Ten Talks 14: Lex in fall of 2020: Oct. - Nov.
#130: Scott Aaronson returns
Simulation
Theories of everything
Consciousness
Roger Penrose on consciousness
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
I admire the fact he was even raising such questions at all. … He is not merely saying quantum mechanics is relevant to consciousness. … much more audacious … even quantum mechanics is not good enough (he is saying) … A quantum computer can be simulated by an ordinary computer; that’s simply not good enough for him. What he wants, is for the brain to be a quantum gravitational computer; He wants the brain to be exploring as-yet-unknown laws of gravity … that would be uncomputable, even if you had an oracle for the halting problem … or as high up you want to go, in the usual hierarchy of uncomptability; he wants to go beyond all of that. Just to be clear, if we’re keeping count of how many speculations, there’s at least five or six of them. First there’s a quantum gravity theory that would support this un-computability. … Most would not support that … Most is consistent with the broad idea of nature being computable … Supposing that he’s right about that, what most physicists would say, whatever new phenomena there are in quantum gravity, they might be relevant at the singularities of black holes, or at the Big Bang … they are plainly not relevant at the brain, which is operating at ordinary temperatures, with ordinary chemistry; the physics underlying the brain, the fundamental physics of it, they’d say we’ve completely known for generations now. Quantum field theory lets us parameterize our ignorance. Sean Carroll has made this case in great detail … QFT screens off … This brings us to the whole idea of effective theories. In the standard model of elementary particles, we have a QFT that seems adequate for all the terrestrial phenomena. The only thing it doesn’t explain are, first, the details of gravita, if it’s probed for the extremes of curvature at very small distance … IT doesn’t explain … very exotic things, far removed from our life on earth. Penrose then needs these phenomena to somehow affect the brain; and then he needs a modification of quantum mechanics; he needs quantum mechanics to actually be wrong. What he wants is what he calls an objective reduction mechanism … once quantum states get large enough, they somehow spontaneously collapse. This is an idea lots of people have explored. There’s something called the GRW proposal that tries to say something along those lines. These are theories that somehow make testable predictions … The very fact that they’re testable … we may in the coming decades be able to test these theories and show they’re wrong. Penrose’s ideas; not the ones about consciousness, but … about the collapse of quantum states. But he’d need more than an objective collapse of quantum states—which would already be (a huge) development … He’d also need for consciousness to somehow be able to influence the direction of the collapse, so that it wouldn’t be completely random; but your dispositions would somehow influence the quantum state to somehow collapse more this way or that way. Finally, Penrose says this somehow has to be true based on an argument based on Goerdel’s Incompleteness Theory. … I don’t think G’s Incomplete-ness Theory can do what he wants it to do here. But that is the tower you have to ascend to to go where Penrose goes. …
L: The I. Theory … says basically there are problems that are not computable
S; No, everyone agrees that … But there are statements; there are true statements of arithmetic … that math will not be able to prove. Famous example; that that system itself is consistent. No formal system can prove its own consistency; that can only be done through a more formal system. Why is that relevant to consciousness? The idea it might have something to do with consciousness is an old one; Goedel himself considered this. … Penrose is updating what they and others had said. The idea that Goedel’s theorem could have something to do with consciousness … IN 1950, when Turing wrote his article about the Turing test; he was referring to that as an old idea, and as a wrong one he wanted to dispense with. .. Penrose and his predecessors wanted to say … that even though this system cannot prove its own consistency; we as humans looking from the outside can somehow see its consistency. The rejoinder to that has been, “Can we really?” Maybe he-Penrose can, but can the rest of us? … Notice that, it is perfectly plausible to imagine a computer … that would not be limited to working within a single formal system. “I am now going to adopt the hypothesis that my formal system is consistent. I’m going to add new axioms to my system.” Goedel would have no problem with this. All it says is, there’s no absolute guarantee that, when the AI adds new axioms, they’ll always be right. … The response is obvious: we don’t have any guarantee that we’re right …
Turing test
L: You took part in the Lubner Prize?
S: No, I didn’t .. a claim was made about a robot
L: You participated as a judge in an exhibition event?
S: Eugene Guzman, that was just me writing a blog post.
L: Did you ever chat about it?
S: I did chat with EG … a bunch of journalists started writing breathless articles about the first chat-bot that passes the Turing test. … they named it Eugene Guzman, it was supposed to simulate a 13-year-old boy … “Well, if you look at Turing’s paper; the percentages he talked about, it seemed like we’re past that thresholds.” I had another way to look at it; let’s try the thing out; see what it does with questions like, “Is Mount Everest bigger than a shoebox?” Answer is, it kind of parrots you, because it doesn’t kow what you’re talking about.
L: You convert the sentences into the meaning of the objects they represent … do common-sense reasoning
S: Right; it wasn’t able to intelligently respond to basic common-sense questions. Let me add—there was a famous … chat-bot in the 60s called Eliza, that managed to actually fool a lot of people. People would pour their hearts out to (her), because she simulated a therapist. This turned out to be incredibly effective; she would parrot back to people what they said. This thing was just a few hundred lines of Lisp code. Not only was it not intelligent; it wasn’t especially sophisticated. Eugene Guzman, from what I could see, was not a significant advance. That was the point I was making. You didn’t need a CS professor to say this. Anyone who was looking at it, and had an ounce of sense, could have said the same thing. Since journalists were calling me; I first said, “I’m a quantum person, not an AI person.” They said—you could try it out.
GPT-3
The world has now seen a chat engine, a text engine. It still does not pass a Turing test; there are no real claims that it passes a Turing test. It comes out of the group from OpenAI. But this, as clearly as Eugene Guzman was not an advance … it’s equally clear that this is a major advance. This is a text engine, that can come up with on-topic, reasonable-sounding completions to just about anything that you ask. You could ask it to write a poem about topic X, in the style of poet Y, and it will make a go at that. Definitely as good, as or better than I would have done. You can ask it to write an essay, a student essay, and it will get something I’m pretty sure would get at least a B- in high school and some college classes. The way it achieves this—Scott Alexander had a wonderful way of putting it. They basically ground up the entire Internet into a slurry. To tell the truth, I’d wondered for a while why nobody had tried that. Why not write a chatbot … by using a corpus consisting of the entire web. Now they finally have done that, and the results are very impressive! People can argue whether it’s a step toward general AI or not. But it’s capability we didn’t have a few years ago. A few years ago, it would have surprised me.
L: Their model, it takes a large part of the Internet, and compresses … without fine-tuning … It does a look-up on the Internet of relevant things. How do you explain it?
S: The training involved … lots of lots of pages of the Internet … lots of electricity, lots of computer power. Cost some tens of millions of dollars.
L: More like four or five million … hope is cost comes down
S: Basically, it is a neural net, what’s now called a deep net; basically the same thing. A form of algorithm people have known about for decades. But it is constantly trying to solve the problem, “Predict the next word.” It’s trying to predict what comes next. It’s not trying to figure out what ought to be true. It’s trying to predict … how whomever wrote up till then … what word comes next.
L:
S: Of course, of course. If there’s a deep question raised by GPT-3, this is it. Is that what I’m doing right now? is that all we do?
L: Is it possible, the intuition people have—this thing will not be able to reason. Do you think it’s possible GPT-5, 6, & 7 will be able to do something … like reasoning.
S: The truth is, we don’t really know what the limits are. GPT-3 was the same as GPT-2 but with a much larger network, more training time, bigger training corpus, and very noticeably better than its immediate predecessor. We don’t know where you hit the ceiling here. That’s the amazing part, and maybe the scary part. At some point, there has to be diminishing returns. It cna’t be that simple, can it? I wish I had more to base that guess on
L: We’re going to hit a limit on the amount of data that’s on the Internet … (57:00)
Universality of computation
Complexity
P vs NP
Complexity of quantum computation
Pandemic
L: So we’re living in the middle of a pandemic … How has your life been changed—how has your perspective of the world changed with this world-changing event …
S: All of our lives have changed … as with no other event since I was born. You would have to go back to WWII … As for how it has changed my worldview; I think the failure of institutions, like the CDC, like other institutions that we sort of thought were trustworthy, like a lot of the media, was staggering, absolutely breathtaking. It was something I would not have predicted. I wrote on my blog that, you know, it’s fascinating to re-watch the movie Contagion from a decade ago, that correctly foresaw so many aspects of what was going on. An airborne virus originates in China, spreads to much of hte world, shuts everything down until a vaccine can be developed; everyone has to stay at home. It gets a lot right. The one thing they could not imagine; in this movie, everyone from the government is hyper-competent, dedicated to the public good—
L: The best of the best
R: Right. There are these conspiracy theorists, who think this is all fake news; there’s not really a pandemic … those are the random people on the Internet that the hyper-competent government people have to oppose. … In trying to envision the worst that could happen … There was a failure of imagination by the movie-maker;s they did not imagine that the conspiracy theorists and incompetents and the nut-cases would have captured our institutions and would be the ones actually running things.
L: So you had—I love competence in all walks of life. I’m so excited by people who do an amazing job. I had a hope that government at its best could be ultra-competent. Two questions. How do you explain the lack of competence; and how can we build a more competent government.
S: Well, there’s an election in two months. It’s not going to fix everything. I have a feeling there’s a ship that is sinking, and it can stop the sinking. There are much much deeper problems. It is plausible to me that much of the failure with the CDC, with some of the other health agencies, even, predate Trump, predate the right-wing populism that has sort of taken over much of the world now. I think that, it is very—I’m actually, I’ve been strongly in favor of rushing vaccines. I thought that we could’ve done human challenge trials, which were not done. We could have had volunteers, to actually get vaccines, get exposed to Covid.
L: So innovative ways of accelerating what we’ve done
S: Yeah—each month that a vaccine is closer, is like trillions of dollars. And of course lives; hundreds of thousands of lives.
L: Are you surprised that it’s taken this long; we still don’t have a plan. There’s still not a feeling that anyone is actually doing anything. There’s a bunch of stuff: there’s vaccines, but there’s also, you could also do a testing infrastructure where everybody’s tested nonstop; contact tracing —
S: I’m as surprised as almost everyone else. I mean this is a historic failure; one of the biggest failures of the 240-year history of the United States … One thing I think has been missing, even from the more-competent side, has been the World War II mentality. The mentality of, let’s just—if by breaking a whole bunch of rules, we can get a vaccine in half the amount of time that we thought .. let’s just do that. We have to weigh …
L: One key aspect to that is deeply important to me; we’ll go to that topic next: the World War II wasn’t just about getting the job done. There was a together-ness to it. I would, If I were president right now: it seems quite elementary to unite the country. Because we’re facing a crisis; it’s easy to make the virus the enemy. And it’s very surprising to me that the division has increased instead of decreased.
S: Well look; it’s been said by others; this is the first time in the country’s history that we have a president that does not even pretend to attempt to unite the country. Lincoln, who fought a Civil War, said he wanted to unite the country. I do worry enormously about what happens if the result of this election are contested? Will there be violence as a result of that? Will we have a clear path of secession? This is all, we’re going to find out the answers to this in two months; if none of that happens, maybe I’ll look foolish. I am willing to go on the record and say, I’m terrified about that.
L: I’ve been reading The Rise and Fall of the 3rd Reich; this is one little voice to put out there. I think November will be a really critical month for people to breathe and put love out there. Do not—anger in that context, no matter who wins, no matter what is said, may destroy our country, may destroy the world. It’s really improtant to be patient, loving, empathetic. It troubles me that even people on the left aren’t able to have a love and respect for people who voted for Trump.
S: I know they are! Because I know some of them. It baffles me, but I know such people.
L: It’s also heartbreaking to me, on the topic of cancel culture. I’ve seen it in the ML community. There is aggressive attacking of people trying to have a nuanced conversation about things. It’s troubling, because nuanced conversation is the only way to talk about difficult topics. When there’s a thought police, speech police, on any nuanced conversation. Like everyone has to, like Animal Farm, chant that racism is bad, and sexism is bad, which is things that everybody belives—they can’t possibly say anything nuanced. It feels like it goes against any kind of progress. You’ve written a little bit about cancel culture. Did you have thoughts?
S: To say that I am opposed to this trend of cancellations, or shouting people down rather than engaging them, would be a massive understatement. I think I’ve put my money where my mouth is. I’ve tried to do something; I’ve defended some unpopular opinions and people on my blog. I’ve tried to defend norms of open discourse; reasoning with our opponents, even when I’ve been shouted down for that. Called a racist and a sexist on social media. I would be perfectly happy, if we had time, to say ten thousand times—of racism, of sexism, of homophobia. What I don’t want to do, is to cede to some particular political faction the right to define what is meant by those terms. Well then you have to agree with all these extremely contentious positions—or else you are a misogynist, or else you are a racist. Don’t people like me also get a say in the discussion, about what is racism, about what’s going to be most effective to combat racism. This cancellation mentality is I think spectacularly ineffective at its own professed goal.
L: What’s a positive way out of it? … I really focus on the positive; I try to put love out there in the world. And still! I get attacked, and I look at that … I haven’t actually said anything difficult and nuanced. You talk about somebody like Steven Pinker. He tries to be thoughtful about difficult topics; and obviously he just gets slaughtered.
S: Well, yes, but it’s also amazing how well Steve has withstood it; he just survived an attempt to cancel him.
L:
S: He’s incredibly unperturbed with this stuff. … My impulse .. is to engage every single anonymous person on Twitter or Reedit … Sometimes that even works. The problem is, there’s the 20,000 other ….
L: Does that ware on you?
S: It does. What is the solution? I don’t know. These problems are maybe harder than P vs. NP. I think there’s a lot of maybe silent support for the open discourse side, the reasonable enlightenment side. That support has to become less silent. A lot of people agree … these cancellations are ridiculous … This faction understands this and exploits it to its great advantage. … Guess what? We’re against racism too, but what you’re doing is ridiculous. The hard part is, it takes a lot of mental energy; it takes a lot of time. Even if you’re … staying on the safe side. The more people speak up, from all political persuasions …
Love
L: Has love played an important role in your life? It’s easy for a world-class computer scientist … you could even call yourself a physicist … to be lost in the books. Has connection for other humans played an important role
S: I love my kids, my wife, my parents. I am probably not different from most people (in that) … Now, I should remind you that I am a theoretical computer scientist; if you’re looking for deep insight for the nature of love, you’re probably looking in the wrong place, to ask me.
L: Is there something to be said about it from the CS perspective?
S: There was this great cartoon; I think it was xkcd. It shows a heart; squaring the heart, taking the four-year transform of the heart, integrating the heart. Then it says, my normal approach is useless here.
#131: Chris Lattner’s 2nd visit, Oct. 2020
Chris Lattner, his second time on the podcast. He’s one of the most brilliant engineers in modern computing. He created LLVM compiler infrastructure project, the Klang compiler, the Swift programming language, a lot of key contributions to TensorFlow and TPUs as part of Google. He’s served as VP of AutoPilot software at Tesla; as a software innovator and leader at Apple; and now is at SiFive as senior VP of platform engineering, looking to revolutionize chip design to make it faster, better, and cheaper.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
As a side note, let me say Chris has been an inspiration to me … he is so damn good as an engineer and leading of engineers; but is able to stay humble … enough to hear voices of disagreement … supported me in the early days of the podcast. Tbh, in the early days, nobody really believed I would amount to much … That’s a lesson for educators. The weird kid in the corner with a dream is someone who might need your love and support in order for that dream to flourish.
Working with Elon Musk, Steve Jobs, Jeff Dean
L: What are the strongest qualities of SJ, EM, and the great and powerful JD, since you’ve worked with each.
C: Easy question (sarcastic). These are three very different people; I guess you could do maybe a pair-wise comparison … if you look at Jobs & Elon; I worked more with Elon. They have commonality; they’re both visionary, very demanding. My sense is Steve is more human factor-focus; Elon’s more tech-focused.
L: What does human-factor mean?
C: Steve’s trying to build things that feel good, that affects people’s lives, how they live. He looks into the future of what people want. I think Elon focuses more on exponentials work, the development of those.
L: Reading biography of Steve … how can a ddesigner talk to engineers and get their respect.
C: I’m not an expert. My sense is he pushed people really hard; when he got an explanation that made sense, he let go. He had a lot of respect for engineering, but he knew when to push. When you know people well, you know when they’re holding back … If you compare the others. Jeff is super-smart; he’s really really nice; well-meaning, the classic Googler, wants people to be happy; combines it with brilliance; so he can pull people together in a really great way. He’s definitely not a CEO type. I don’t think he’d want to be. … Really hard to compare him to either of those two. I think Jeff leads through tech and building it himself … and inspiring them. But each of these people, with their pros and cons, are inspirational and have achieved amazing things. I’ve been very fortunate …
L: Something you’ve picked up. … about how to lead?
C: … you really need to know what you’re talking about. Being grounded on the mission. Understanding what people are looking for; why they’re there. With Tesla, the unifying vision. People are there because they believe in … Also understanding what really motivates people. How to build a plan … that can be executed. Depends on a lot of things …
L: Two interesting things you mentioned. That you have to know what your’e talking about. You’ve worked on a lot of really challenging technical things. Assuming that you weren’t born with that tech savvy; how do you master that tech expertise. Even at Google … very very challenging stuff.
C: I’ve become more comfortable, with experience, with not knowing. A major part of leadership is not about having the right answer, but getting the right answer. If you’re on a team of amazing people, it’s a question of how you get people together … How you get people to be vulnerable … it’s the start of something beautiful. How you provide an environment … it’s not top-down, forcing stuff.
L: So asking dumb questions.
C: Yeah, dumb questions are my specialty. I’ve been in the hardware realm recently; I don’t know a lot of this … It turns out if you ask a lot of dumb questions, you get smarter really quick.
Why do programming languages matter?
Python vs. Swift
Design decisions
Types
Programming languages are a bicycle for the mind
Picking what language to learn
L: A small tangent. I’ve been recently eating only meat. Most people think that’s horribly unhealthy. Whatever the science is, it just doesn’t sound right
C: Back when I was in college; we did the Atkins diet, that was a thing.
L: We have to always give these things a chance. If I eat meat, I can be more focused than usual; I just feel great. Python is similar in that sense for me.
C: Where are you going with this?! (laughter)
L: I had a stupid smile on my face when I first started using Python. I would see the world. I would be empowered to write a script to do some basic … things on my computer. Pearl didn’t do that for me.
C: None of these are about which is best; but there’s definitely better and worse here. If you look at Pearls, you get bogged down in tables versus arrays … let’s not do this
L: Can I create systems for myself that empower me to debug quickly. …
C: You can think of types in a programming language as being kind of a cert. … How do people learn new things? People don’t like to change; people don’t like change around them either. We’re all very slow to adapt and change; usually a catalyst is required … Learning programming is akin to finding an excuse. Build a thing that … the ecosystem’s ready for. … you would use Swift for that.
L: Android!
C: Swift runs on Android. Swift runs on LLVM; LLVM runs everywhere; it builds the Android kernel … it runs on Windows …
L: Swift UI; there’s a thing called UI kit. Can you build an app with Swift?
C: Well Swift UI and UI Kit are Apple technologies … you go to Android and you don’t have that library. They have a different ecosystem of things. Building a UI with Swift on Android is not … great … right now.
L: If I wanted to learn Swift … one example of that is Swift for Tensor Flow … With TEnsor Flow and PyTorch, it’s beautiful how quickly the community can switch between different libraries … TensorFlow has been stepping up its game. … It gets people excited; people go, “I have to learn this.”
C: There has to be a reason; a catalyst. … One of the fun things about learning programming language. Even Lisp, I don’t know. When you start doing it, you start learning new things. It forces you to explore; it puts you in learning mode …
L: … I wish there was a window …
Most beautiful feature of a programming language
L: Before I ask that; I remember on Python I saw “Lisp comprehensions.” It was fun to do that kind of … something about it, to be able to filter through a list; to create a new list all on a single line. IT made me fall in love with the language. What, to you, is the most beautiful feature in a programming language, that you’ve ever encountered. In Swift, and then outside of Swift.
C: The thing you have to think about—what is the goal? You’re trying to help people get things done quickly. You need libraries. High-quality ones. And a user-base. What enables high quality libraries? Languages that enable high-quality libraries, versus … … What I mean is, expressive libraries that feel like a natural integrated part of the language itself. … In Swift, “int” and “float” are all part of the library. Int is not hard-coded into Swift. Because int is just a … Well, hopefully you do like “Int.” Any language features you needed, you can also find in your own types. If you needed a quaternion, it doesn’t come in the standard … There’s a specialized group of people who care about them; but those people are also important. It’s not classism. So those people can build something … that feel native, that feel like they’re built into the compiler …
L: What does it mean for the int to not be hard-coded in. What is an “int”?
C: An integer, a 64-bit integer. …. If you look at how it’s implemented, it’s just a struct in swift. How do you add two structs? Well, you define plus. You can define it on int, or on your own struct. You can add it as a method.
L: So you define operators. That, to you, is beautiful. When you enable others to create libraries that are not “hacky.”
C: Right, they feel native. One of the best examples of this is Lisp. All the libraries are basically part of the language.
L: Can you give a counterexample …
C: I’ll give two examples, Java and C. They both allow you to define your own types. But int is hard-coded in the language. Why? Java … Int gets passed around by value, but if you make like a pair, or something, a complex number, it’s a class in Java; and it gets passed by reference, by pointer. And so you lose value semantics; you lost math. That’s not great; if you can do something with int, why can’t you do it with my type. That’s not great … you as a user of the language (should) have almost as much power as the people who designed the language.
L: I’ve gotten used to that. .. the old first-principles thinking. Why are we doing it this way?
Walrus operator
L: It hit me. … why are we using the equals sign for assignment?
C: It’s wrong. In Pascal, they use colon-equals for assignment.
L: SQL … but how do you then decide to break convention. To say, you know what, everybody’s doing it wrong. We’re going to do it right.
C: It’s like ROI. … You’d say, cool—this is theoretically better; it’s better in which ways? Do I define a new class of bugs? … x = y … turns out you can solve that in lots of ways. Clang, JCC, turns out they detect it as a likely bug …
L: One of the important things about programming language design is, you’re literally creating suffering in the world. … You can think of it as, minimizing suffering.
C: Let’s come back to that. … There’s a lot of detail that goes into these things. Equal, in C, returns a value. That’s messed up; it allows you to say x = y = z ..
L: Is that messed up?
C: What I mean is, it is very rarely used for good, and it’s often used for bugs.. ..
L: That’s a good definition for “messed up.”
C: One thing about Swift, a reason it’s actually good. When we launched it, we announced, people could use it. … We said hey it’s open source; there’s this process by which … people can help shape the language. As part of that process, a lot of really bad mistakes got taken out. C used to have the ++ and minus-minus operators. That got cargo-culted from C early on. … That means brought forward without really considering it. Not the most PC term. Got pulled in without very much consideration. It got ripped out because they lead to confusion; the have very little value. x += 1 turns out to better … it makes more sense.
L: This podcast is about drama. You mentioned Pascal and :=; there’s something called the Walrus operator, added in Python 3.8. It’s interesting not just because … as a feature, it returns the value of the assignment. But on the other side of it, it’s also; it toppled a dictator. It drove Guido to step down … the toxicity of the community. What do you think about the Walrus operator? Is there an equivalent in Swift that really stress-tested the community?
C: If I look past its details; one thing that makes it polarizing is that it’s syntactic sugar.
L: I’ll play devil’s advocate. Is that objective or subjective? Can you argue that basically anything is syntactic sugar?
C: No. For example, the type system; can you have classes versus … Do you have types or not? One type versus many types … doesn’t affect .. sugar. So having types is not about sugar-ing it. Swift has sugar. It has “if-let” and various operators meant to concise-ify various use cases. Problem with syntactic sugar; you have this horrible tradeoff that becomes completely subjective. How often does this happen and does it matter? One thing about human psychology—people overestimate the burden of learning something. If it was there from the beginning, of course, it’s just part of Python. … Now with Guido, I don’t know him well. .. I’ve met him a couple of times; the sense that I got out of that whole dynamic; he had put not only the decision-maker weight on his shoulders, but it was tied to his personal identity and he took it personally.
L: And that can wear you down?
C: People say, Guido, you’re a horrible person, I hate this thing, bla bla bla. Sure it’s 1% of the whole community, but it’s a big community, and then that’s a lot of hate mail.
L: Looking just at the Python core developers. … It seemed like the majority, on a vote, were opposed to it.
C: I’m not that familiar with it.
L: The majority … It’s not that they were against it; they just were unpersuaded it was a good idea. That results in … a stalling feeling. Now, from my perspective. It’s interesting, if you look at politics, the way Congress works: it slow down everything.
C: It’s a dampener.
L: But that’s dangerous, because … if it dampens things
C: Well .. it’s a low-pass filter, but if you need billions of dollars injected into the economy, then suddenly stuff happens. I’m not defending our political situation.
L: You’re talking about a global pandemic … I was hoping we could fix the health care system
C: I’m not a politics person. I don’t know. When it comes to languages, the community’s kind of right: it’s a very high burden to add something to a language. You have a community of people building on it, and you can’t remove it. If there’s a community of people uncomfortable with it, then taking it slow is an important thing to do … and there’s no rush. Particularly with something that’s only 25 years old, and coming into its own.
L: What about features?
C: I think the issue with Guido; maybe he realized the language had outgrown him. Python isn’t about Guido anymore; it’s about the users…. the users own it. He’s spent years of his life on Python. … He imagines, this is my thing. But you can also understand the users—they feel like, this is my thing. I use this.
L: If we could talk about leadership in this. … Someone should make a Walrus operator shirt. It represents to me … the burden of leadership. To push back, I feel like most difficult decisions, just like you said, there will be … divisive-ness over. Seems like leaders need to take those risky decisions. If you listen … with some probability it will be the wrong decisions.
C: Amazing founders … they understand what’s happened, how the company got there. We’ve been doing thing X for 20 years, but today we’re going to do thing Y. … They make a major pivot for the whole company, and it’s the right thing. When the founder dies, the successor doesn’t feel … ready to make that same type of decision. There’s many reasons for that. One is, they weren’t there for all the decisions that were made, and so they don’t know the principles in which those decisions were made. … Once the principles change, you should be obligated to change what you’re doing, and change direction. If you don’t know how you got to where you are, it seems like gospel, and you aren’t going to question it.
L: That’s so brilliant. It’s so much higher-burden, when as a leader you step into a thing that’s already worked for a long time. …
C: Even if you decide to make a change, even if you’re theoretically in charge … you’re just a person! You have to motivate the troops, get them to buy into it and believe in it. Else, they’re not going to be able to make the turn. There’s only so much power you have as a leader.
L: Are you still BDFL … for LLVM?
LLVM
L: What’s the role? On Swift you said there’s a group of people.
C: Everybody on the core team takes their role really seriously; we all really care about where Swift goes. You’re almost delegating final decision making to the wisdom of the group. Yes, in the community, some people are very annoyed …
MLIR compiler, framework
SiFive semiconductor design
Moore’s Law
Parallelization
Swift concurrency manifesto
Running a neural network fast
Is the universe a quantum computer?
Effects of the pandemic on society
L: .. if we step back and zoom out, about the way you did that work, and we look at Covid times, this pandemic we’re living through … if I look at the way … people are talking about, this might last for a long time. Not just the virus, but the remote nature …
C: The economic impact. Yeah, it’s gonna be a mess!
L: What’s your prediction? From SiFive, to Google, to all the places you worked … how is this whole place gonna change.
C: I can only speak to the tech perspective … I think its’ going to be really interesting. the Zoom culture of being remote, on videochat all the time, has intereesting effects on time. It’s a great normalizer that will heelp communities of people that have traditionally been underrepresented. You can have conversations without personal appearance being part of the dynamic. .. Remote employees are now on the same level and footing as everybody else—nobody has whiteboards! You’re forcing people to think asynchronously, because it’s harder to get people physically together. Finding new ways to solve problems, whi. .. OTOH, it just sucks! Right?
L: The actual communicsation, or it sucks not being with people on a daily basis and collaborating
C: All of that, everything, this whole situation. Primarily, I think most humans like working physically with humans. Not everybody, but many people are programmed to … we get something out of that, that’s very hard to express, at least for me. The question to me is, when you get through that time of adaptation. You get out of March April, and into December
L: It’s already terrifying
C: You think about, what is the nature of work? Humans are adaptable … What is it that comes out of this? Are we better or worse off? You look at the Bay Area, housing prices are insane. Because if you don’t have proximity, you end up paying for it in commute. There’s big social pressure you will be there for the meeting. That will be way better … more common to have remote employees …
L: Do you know people moving?
C: One family friend moved back to Michigan … they were in tech.
L: Friends of mine, are in the process of or have already lost their business. The thing that represents their passion … people who run gyms. Also … people look themselves in the mirror and ask themselves, what do I want to do in life? … That results, moving, lleaving the company you’re with. Do you think we’re going to see that a lot?
C: I can’t speak to that. We’ll see it at a higher frequency …. I think what you’re saying is, big life decisions that you make yourself; then there’s decisions that are made for you. “I got fired,” you’re forced to act. Those latter kind of moments, when a global pandemic comes and wipes out the economy; I think that does lead to more reflection. You’re less anchored on what you have; it’s more of a fresh slate—cool, I could do anything now. Do I want to do the same thing I was doing? Is it time to go back to college? Learn a new skill? Time to spend time with family? Literally move in with parents? Things that weren’t normative before; the value systems change. I think that’s a good thing, in the short-term at least. Tehre’s been an over-optimization along one set of priorities for the world. … I think it could be good; there could be more innovation that comes out of it.
L: What do you think about all the social chaos in the middle of it. Do you think it’s all gonna be okay?
C: It sucks … I don’t think all the humans are going to kill all the humans. I look at it as, progress requires … a catalyst. You need a reason for people to be willing to do things that are uncomfortable. I think the U.S. and the world is an un-optimal place to live in for a lot of people. We’re seeing a lot of unhappiness; all the pressure in the world is igniting debate that should’ve happened a long time ago. Offline you were talking about politics, and wouldn’t it be nice if politics moved faster … Well. People are inherently conservative. Particularly people who have heavy burdens on their shoulders because they represent thousands of people. OTOH, when you need change, how will you get it? Well, a global pandemic probably will lead to some change.
L: Let me know if you’ve observed this as well. In the ML community, and I’m guessing it’s prevalent in other places. In 2020—an increased level of toxicity. People are quicker to pile on; are just harsh on each other, to just mob and pick a person that screwed up, and make it a big thing. Is there something that we can, have you observed that in other places?
C: I think there’s an inherent thing in humanity, an us-versus-them thing. You want to succeed, as meausred by, it’s realtive to somebody else. With the Internet, the world’s getting smaller, and some of the social ties—my town versus your town’s football team, are turning into larger, and yet shallower problems. People don’t have time; clickbait, and all these things feed into this machine, and I don’t know where that goes.
L: I mentioned to you offline; I’ve had a few difficult conversations scheduled; some of them political-related. Some within the community. One of them I’ve talked to before, and will again, is Yann LeCun, who got crap on Twitter for talking about a particular paper and bias within a data set. There’s been an irrational overexaggerated pile-on on his comments, because he made pretty basic comments about the fact that if there’s bias in the data, there’s going to be bias in the results. People piled on, because they said he trivialized the problem of bias. That’s a good point, but that’s not what he was saying, and the implied response that he’s basically sexist and racist is something that completely drives away the possibility of nuanced discussion. One nice thing about long-form conversation is that you can talk it out; you can lay your reasoning out. Even if you were wrong, you can still show you’re a good human being underneath it.
C: … productive discussion. How do you get to the point where people can learn and engage? Rather than a shallow “like” and keep moving … You’d see that as reinforcing individual circles and the us-versus-them thing.
L: Yeah. The people that bother me most on Twitter .. is not the people who get emotional and angry. It’s the people who prop them up. It’s … I think what we should teach each other is to be sort of empathetic.
C: It’s easy to forget … that sometimes people just have a bad day. I’ve been in the situation where, between meetings, I fire off a quick response to an email .. phrased it wrong, and now it’s something that sticks with people. It’s not because they’re bad, or you’re bad. You didn’t mean it that way, but it really impacted somebody … the way they interpreted it. I have a lot of otpimism in the very-long-term, but it’s always a rough ride. You came into this … asking what does Covid mean? I think it’s very bad in the short-term, but I think it’ll lead to progress, and for that I’m very thankful.
L: Bad in the short-term
C: Yeah! People out of jobs … the question is, when you look back 10, 20 100 years from now, how do you evaluate the decisions that are being made right now? You ingtegrate across all the short-term horrible-ness … is the improvement across the world or the regression across the world significant enough to make it a good or bad thing.
L: Yeah. For that, it’s good to study history. Hard for me right now because I’m reading The Rise and Fall of the Third Reich. … The thing that worries me the most … is the pain that people feel when a few things combine: economic depression, and just being disrespected in some kind of way. The German people were really disrespected by most of the world …. Then all you need is a charismatic leader to either go positive or negative, and both work.
C: It takes advantage of that inflection point that the world’s in; they can take advantage of it, either way, good or bad.
L: When history’s written … probably history’s going to remember 2020, either for good or bad. It’s up to us to write it so it’s good.
C: I’ve observed … most people act as if the world doesn’t change. Knowingly, you make a decision where you’re predicting the future based on what you’ve seen in the recent past … OTOH, the world changes. All the time! Constantly, for better and worse. If you’re interested in something that’s not right. What is the inflection point that led to a change? The catalyst that led to the explosion that led to that bill … Maybe if you pull together the right people, the right ideas, maybe you can drive that change … in a way that hurts fewer people. Often it’s a combination of multiple factors.
L: That’s the optimistic view.
C: Human nature, we can look to all the negative things that humanity has … the cruelty, the biases, the … humans can be horrible. OTOH, we’re capable of amazing things. (Laughs) The progress across 100-year chunks is striking! …
L: God, the stuff we did in the last 100 years is unbelievable. It’s kind of scary. Scary exciting; it’s kind of sad, the kind of technology that’s gonna come out; we’re probably too old to really appreciate. It’ll be like, kids these days with their VR, and their TikToks …
GPT-3
L: … these language models … What’s your thoughts on this exciting world; it connects to computation, actually. Language models that are huge; that use many, many computers…
C: It depends on what you’re speaking to, there. I think there’s been a pretty well-understood … maxim in DL that if you make the model bigger … assuming, you train it right, that' you’ll get a better model. So GPT-3 wasn’t that surprising, OTOH, a tremendous amount of engineering made it possible. … When GPT-2 came out, there was a provocative blog post—we’re not going to release it becasue of the social damage it could cause, if misused. That’s still an issue. … I think GPT-3 is a huge technical achievement. What will GPT-4 be? …
L: … distributive computing. Is there technical challenges that are interesting? In terms of a piece of code that. .. GPT-4 … might have, IDK, hundreds of trillions of parameters, running on thousands of computers. …
C: Well, today, you can write a chck and get access to a thousand TPU cores .. in Google Cloud … I think it’s a question about utility (not scale) … And when I look at. the transformer series of architectuers that the GPT series is based on; … they’re actually simple designs … the train measurements … … … … … … …
Software 2.0
L: On the data side … if you can think of data selection as a type of programming. Karpathy talked about Software 2.0 …
C: Let me try to summarize Andre’s position before I disagree with it. Andre Karpathy is amazing; an amazing engineer
L: Also a good blog-post writer
C; He’s also really sweet. His premise is software is suboptimal; we can all agree. He also surmises DL is really great, because you can solve problems … with less ad hoc code … that might not work. If you start replacing systems of imperative code with learning models, you get a better result, ok. He argues Software 2.0 is a learned set of models, and you get away from writing code. He’s given talks about swapping more and more parts of code to being learned … If you’re predisposed to liking ML, that’s a good thing. This is also good for accessibility in many ways; certain people aren’t going to write C code, so having a data-driven approach could be valuable. OTOH there are huge tradeoffs. I don’t think Software 2.0 is the answer. I don’t think it’s a replacement for Software 1.0; I think it’s a new programming paradigm. … Structured programming … or functional programming from Lisp … or object-oriented programming, or generic programming, with code reuse … or differentiable programming … which I’m excited about. … ML is amazing at solving certain types of problems. You’re not going to write a cat detector, by writing C code. ML is absolutely the right way to do it. I would say learn models are the right way to work with the human world … … … humans are very very difficult to characterize. OTOH, imperative code is too. DL models are hardware intensive, energy intensive. You can provably implement any function with a learned model, but that doesn’t make it efficient. If you care about energy usage, it’s useful to have other tools in the toolbox. … .Problems with data and bias of data … … …
L: At the same time, the boot loader that works for us humans looks a lot like a neural network. .. You can cut out parts of the brain … neuroplasticity. Interesting, how much of the world’s programming could be replaced by Software 2.0
C: It’s provably true you can replace all of it. It’s not a question of “if.” It’s, an economic question. What kind of talent can you get? What kind of data can you collect? One of the reasons I’m interested in ML as a programming paradigm … we’ve seen being laser-focused on one paradigm, puts you in a box, that’s not super-great. Object-oriented programming was THE fad in the late 80’s. But functional programming is good, too—people forgot! … It’s not about whose tribe should win. That shouldn’t be the question. It should be how do you make it so people can solve these problems the fastest. You look at Reinforcement Learning as one subdomain of this. It requires the integration of a learned model, combined with your Atari, with the robot control for the arm, right? So it’s not just about one paradigm; it’s integrating that with all the other systems you have.. … To me, the interesting thing is, how do you get the best out of this domain, and help people achieve things …
L: Ok, but do you think it’s possible these language models … could replace some aspect of compilation, or do program synthesis …
C: Yeah, absolutely. They’re extremely powerful; I think people underestimate them.
L: I have access to GPT-3 API; would I be able to generate Swift code?
C: It’s probably not trained on the right corpus. It could probably generate some Swift, but not enough to be useful. But … if you were in a position to train GPT-3 … it could do that. The question is, how do you express the intent of what you want filled in? It’ll put in some for loops, some classes. … There’s an unsolved question—how do I express the intent of what I want to fill in? … I don’t know that these models are up to the task, if you say—here’s the scaffolding; and here’s the assertions at the end. You want a generative model, yes, but you want some loop-back …
L: So, it would generate not only a bunch of the code, but also the checks, the tests, that do the testing.
C: I think the humans would generate the tests. The tests are the requirements. You have to express to the model … you don’t just want to generate code
L: … compelling idea that GPT-4 model could generate checks that are more high-fidelity, that check for correctness. … Say I ask it to generate a function that gives me the Fibonacci sequence
C: So decompose the problem. You want syntactically correct Swift code that’s interesting. But then you need to add the requirements. The human needs to express that goal …
L: It can generate stuff. .. It can generate HTML; it can generate basic for loops
C: But how do you say, “I want google.com” or “I want a web page that’s got a shopping cart.”
L: Have you seen these demonstrations: you say “I want a red button, with text that says hello.” It generates the correct HTML. You have to prompt it with similar kinds of mappings … quite impressive; again, that’s very basic. The HTML is kind of messy and bad. But the idea, is the intent is specified in natural language
C: That’s really cool.
L: The question is the correctness of that. You can see, “Oh the button’s red.” But for more complicated functions where the correctness is harder to check. … A giant thing that does some kind of calculation. It’s interesting to think—should it try to …
C: This is way beyond my experience. The thing I think about is, there doesn’t seem to be a lot of equational reasoning. … … You need higher-level logic. You can talk to Yann about that; see what the bright minds are thinking about right now. It’s still really cool …
L: Who knows; maybe reasoning …
C: … is overrated! Do we reason? How can we tell? Are we just pattern-observing, and reverse-justifying to ourselves?
L: What it needs is to be able to tell stories to themselves
C: That’s what humans do. We don’t know why we make decisions. We have this thing called intuition … is that science? Not really.
Advice for young people
Meaning of life
#132: Hacker George Hotz (2nd visit)
Hotz’ first appearance on the podcast is transcribed here.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Will human civilization destroy itself?
Where are the aliens?
Tic Tac UFO and Bob Lazar
Conspiracy theories
The programming language of life
The games that humans play
L: … brilliant people like you that really ponder living a long time … I don’t think you’re really considering how meaningless life becomes
G: I wanna know everything, and then I’m ready to die.
L: Isn’t it possible, the reason you want to learn “everything” is because you don’t have time to learn everything.
G: (repeats his previous statement). It’s a value not in service of anything else.
L: I’m conscious of the possibility that the engine of curiosity you’re speaking to, is actually a symptom of the finite-ness of life. Like, without that finite-ness, your curiosity would vanish, like a morning fog.
G: Okay, cool. I’ll go solve immortality but cure the thing in my brain that tells me I’m immortal …
L: Oh, like changing the code.
G: If that’s the problem … I’d like to have backup copies of myself; which I check in with occasionally to make sure they’re okay with the trajectory
L: But sometimes it’s not reversible … Once you figure out the cheat code … it ruins the game for you.
G: I know the feeling. But that just means our brain manipulation code isn’t good enough yet.
L: It’s possible if we figure out immortality, all of us will kill ourselves.
G: I’m not killing myself till I know everything.
L: that’s what you say now, because your life is finite!
… L: That talk . .. is not necessarily referring to a simulation; and our universe as a program running into a computer. Do you also think about the potential of a Bostrum … Elon Musk … they talk about an actual computer that simulates our universe?
G: I don’t doubt that we’re in one … if totally unfalsifiable, then who cares?
L: What does that experiment look like? … … … The step that precedes the step of getting power from this knowledge … … you can discover and exploit cracks in the simulation, or in the physics of the universe.
Memory leaks in the simulation
G: I mean, like a memory leak would be cool. … some scrying technology .. the parallel ability, like remote viewing; being able to see somewhere where you’re not. … It’s a memory leak, basically.
L: Yeah, you’re able to access parts you’re not supposed to, and thereby discover a shortcut.
G: The ability to read arbitrary memory. …
L: Oh boy, the writing is a big, big problem. The moment you can write anything, even if it’s just random noise—that’s terrifying.
G: Even without that, the nanotech stuff that’s coming, that’s …
Theories of everything
L: Eric Weinstein’s been working on a TOE called Geometric Unity. Weinstein’s TOE is very beautiful. Have you paid attention to any of that?
G: What would make me pay attention is … make a test-able prediction. I’m not interested in theories for their intrinsic beauty. I’m interested in theories that give me power over the universe
L: That’s beuatiful … physicists say they’re interested, and they fail to add “that give me power in the universe.”
G: I want power!
L: (laughs) That’s true and that’s where people … are worried, from aliens, from governments—who owns that power? Is it a George Hotz; is it thousands of hackers across the world? Is it Mark Zuckerberg? I don’t know if anyone trusts any one individual with power …
G: That’s the beauty of block chains.
L: Which we’ll talk about. On Twitter, a bunch of people pointed me to a story where you went into a restaurant in NY, and ran into a bunch of folks from a crypto company, that are trying to scale up Ethereum … they had a technical problem. … They recognized you; explained their problem. You stepped in and helped them solve it … legend status story. Can you tell me this story?
G: It’s a true story … a flattering account of it. The company’s called Optimism, a spinoff of Plasma; they’re trying to build L2 solutions on Ethereum. Right now every Ethereum node has to run every transaction on the Etheereum network. …. … This doesn’t scale. You get more block-chain security, but BC’s already so security; can you trade some of that off for speed. … They built … a kind of sandbox for Ethereum …
L: What’s L2
G: L2 is layer 2. L1 is the base ethereum chain. L2 is a computational layer that runs elsewhere but is still secure by Layer 1.
L: And Ethereum is a popular cryptocurrency. And a lot of interesting technological innovations there. Maybe you could also slip in anything that’s exciting to you in the Ethereum space.
G: Well bitcoin is not ___ complete. Ethereum is not totally ___ complete … except for the gas limit.
L: What’s the gas limit?
G: That’s Ethereum’s word … you buy gas with ether, to keep people from ___ the network.
L: Bitcoin’s proof of work …
G: So is Ethereum
L: So we’re all talking about Ethereum 1.0. What were they doing to try to scale …
G: They said if we could run contracts elsewhere and only run the results of the computation. Problem is, somebody could lie about what the results are. You need a resolution mechanism. That could be expensive. … Someone staking $10K on their statement of what happened … you can secure using those kinds of systems. It’s effectively a sandbox which runs contracts.
Ethereum startup story
G: You have to replace Cis-calls with the Hypervisor. …
L: What are those?
G: The Chrome process that’s doing a rendering … if it tries to make an Open-Cis call in Linux … you have to request … through hyper-visor process … I don’t know what it’s called in Chrome, “Can you open this file for me?” …
L: So what are the boundaries of the sandbox?
G: One of the calls … writing any state to the Ethereum block chain … writing state is one of those calls you’ll have to sandbox in Layer 2, because if you write Layer 2 arbitrarily write to the Ethereum block chain
L: Layer 2 is tied to Layer 1 …
G: You have to replace a bunch of calls with calls into the hyper-visor. Instead of doing the Cis-call directly, you replace it with call to the Hv. … So solidity is the language most ethereum contracts are written in. It compiles to a bite code. Then they wrote this thing called the transpiler … This transpiler was a 3,000-line mess. And it’s hard to do. You have to deconstruct the bite code, change things about it, and reconstruct it. So as soon as I hear this, I just think, “why not just change the compiler.” I asked how much they wanted it, in dollars. And then … I wrote the compiler.
L: (laughs) Cute is a good word for it. And it’s C++ …
L: So when asked how you were able to do it, you said you just gotta think, and then do it right. Can you talk about that?
G: It doesn’t really mean anything
L: No deep insights
G: I say, do you want to be a good programmer? Do it for 20 years
Cryptocurrency
L: Yeah; there’s no shortcuts. What are your thoughts on crypto in general
G: I’m bullish. … Two ideas. One, the ___ consensus algorithm; this idea people can reach a group consensus using a straight-forward algorithm, that’s wild. ___ Nakamoto. You ask me who I look up to; the answer is, whoever that is.
L: Who do you think it is? Elon Musk? You?
G: I doubt it’s EM. Definitely not me. … Reaching consensus … is an extremely powerful idea from crypto. The second idea is this idea of smart contracts. When you write a contract between two parties. When there are disputes, it’s interpreted by lawyers. Let’s compare a lawyer to Python.
L: That’s hilarious!
G: Python, I’m paying 10 cents an hour; lawyers cost $1,000 an hour. So Python is 10000x better on that axis. Lawyers don’t always return the same answer. Python almost always does. … If you can make smart contracts; this whole concept of code is law, I love. I would love to live in a world where everyone accepted that fact.
L: Talk about what smart contracts are.
G: Think of something as simple as a safety deposit box. … One that holds a million … I have a contract with the bank that says, 2 of these 3 parties must be present to .. get the money out. That contract is only as good as the bank and the lawyers. Say somebody dies, and there’s a big legal dispute. It’s so messy, and the cost to determine truth is so expensive. Versus a smart contract, which just uses cryptography to see if two out of three keys are present. So then I can have certainty; that’s all businesses want is certainty. …
L: Yeah I wonder how many agreements in this world. … we’re talking about financial transactions only, right
G: You could go to anything. You could put prenup in the block chain. Sorry divorce lawyers, you’re going to be replaced by Python!
L: That’s another beautiful idea. … If you look 10-20, 50 years down the line. .. do you see any of the cryptocurrencies winning out. What’s your intuition about the space?
G: I don’t speculate; I don’t care which one of these projects wins. I’m in the bitcion is a meme-coin camp. Why does it have value? … When I found out what the block-size debate was …
L: What is it?
G: It’s too stupid to even talk about. The governance of Ethereum seems much better. I’ve come around on proof-of-stake ideas. Very smart people thinking about some things.
L: Governance is interesting. It does feel like, even in these distributive systems, leaders are helpful (like Vitalek). They put a face to a project.
G: Well, geniuses are helpful—like Vitalek. …
L: So you think the reason he’s the face of Ethereum is ‘cause he’s a genius. That’s interesting. It’s interesting to think about that we need to create systems in which … the “leaders” that emerge are the geniuses in the system. That’s arguably why the current state of democracy is broken, is the people emerging as leaders are not superstars of the system. … At least for now, in the crypto world, the leaders are the superstars.
G: Imagine if at the debate, they asked “What is the sixth amendment. What is the integral of e^x. What’s Bayes rule?” It would be interesting to see what they answered.
L: My standard was even lower; I would’ve loved to see … I’ve talked to historians. They don’t have a PhD .. like a Dan Carlin type character. And it’s, holy shit, how did all this information get into your head? They connect Genghis Khan to all of the 20th century; they know every single battle that happened …. the individuals, all the documents involved; they integrate that knowledge into their everyday life. That’s what competence looks like. I’ve seen it with programmers too; that’s what great programmers do … it’ll be really unfortunate those people aren’t emerging as our leaders …. For now, at least, in the crypto world, that seems to be the case. You can imagine that in 100 years, that’s not the case (in crypto)
G: The crypto world has one big thing going for it and that’s the idea of forks … I mean, you know … imagine … a less controversial example—this was in my joke app in 2012. Barack Obama? Mitt Romney? Let’s let ‘em both be president. Imagine we could fork America. People could invest in one, put it in the other. You have this in crypto. Ethereum forks into Ethereum, and Ethereum Classic. Companies should be able to fork. I’d like to fork Nvidia.
L: Yeah, business strategies. Try them out and see what works. Take comma.ai that closes its source and one that’s open source, and see if it works. Try one that is purchased by GM, and another …
G: The beauty of comma.ai is that someone could actually do that.
Self-help advice
L: … How do you keep learning new things? Any practical advice—if you were to introspect. Taking notes? Allocating time? Or do you mess around.
G: I’ll write a self-help book and charge $67 … I’ll write that all this advice is completely meaningless. The big lesson—I can’t make a meaningful answer to that.
L: Let me translate that: You haven’t thought about what it is you do, systematically. This is really clear with athletes. Some are the best in the world at something, and they have zero interest in writing a self-help book. But some athletes become great coaches and they love the analysis, the overanalysis. Perhaps at your age; you’re in the middle of the battle; like the warriors
G: This is a fair point. I do have a certain aversion to this deliberate, intentional way of living life.
L: Eventually … the hilarity of this, because this is recorded—it will reveal the absurdity when you eventually do publish this book. “The Story of Comma Ai” …
G: You might be able to learn some cute lessons … but if you’re asking generic questions like, “How do I be good at things?” … I don’t know—do them a lot!
L: But … learning things outside your current trajectory. … it is interesting; you become really busy; you’re running a company. (George laughs) … There’s natural inclination and trend; the momentum of life carries you into a direction of wanting to focus; it gets harder and harder with time. You get really good at certain things. … You do this because your life … streams. You’re on the fly figuring things out; you don’t mind looking dumb.
G: Sometimes I try things and I don’t figure them out. My chess rating is 1400; it’s pathetic. I know I could do it better … “Don’t play 5-minute games; play 15-minute games,” I know these things; it doesn’t fit nicely in my knowledge trey.
G: Solve self-driving cars while delivering shipp-able intermediaries …
L:
G: I can talk about what solve self-driving cars means. It means, you’re not building a new car; you’re building a person replacement … someone who can drive you with a better-than-human level …
L: And delivering shippable …
G: It’s a way to keep us honest. If you don’t have that; it’s easy, with this technology, to think you’re making progress when you’re not. You can set any arbitrary milestone, meet it, and still be infinitely far away from solving self-driving cars.
L: So it’s hard to have real deadlines … when you don’t have revenue. Is revenue essentially a thing we’re talking about
G: Capitalism is based around consent. The way you get revenue … real capitalism is based around idea that if we’re getting revenue, we’re providing at least that much value for another person. We’re providing that value or you wouldn’t buy it.
L: Whirlwind tour of products comma.ai provides …
G: The only product we sell today is the Comma 2
Comma two
L: Which is a piece of hardware with cameras …
G: You can think of it as kind of like a person. It has eyes, ears, a mouth, a brain … and a way to interface with a car. …
L: … what’s the sexiest thing about Comma 2 feature-wise
G: Our slogan summarizes it well: “Make driving chill.” … If you like cruise control, imagine that but much much more
L: So adaptive cruise control things like slow down for cars in front of it; also doing lane-keeping, better and better over time. ML-based. There’s cameras; a driver-facing camera, too. What else is there? The hardware versus software. Open Pilot versus the actual hardware of the device. Can you … that distinction?
G: …
L: By cell phone you mean … Snapdragon
G: Yes. .. It has a wi-fi radio, it has a screen. We use every part of the cell phone
L: And the interface to the car is specific to the car … you support more and more cars
G: … … Cars are actually surprisingly similar
L: Can is the prototype on which cars communicate. So what’s the software side. What’s open pilot?
G: More complicated … Well, so you have an ML model … it’s a blob, of weights. “Oh, it’s closed source!” It’s a blog of weights; what do you expect.
L: NN-based?
G: It’s the software around that NN . .. if given the software where you want to send the car …
L: …
G: It runs the sensors, does a bunch of calibration for the neural network. deals with, if the car’s on a banked road, you have to counter-steer against that.
L: …. … How has Open Pilot improved?
G: I promised you that you’d be able to remove the Lains … the lateral policy is now .. end-to-end, You can turn lanes off and it’ll drive well, on user data. This year we hope to do the same on longitudinal policy.
L: You don’t appear to be doing lane detection or lane-marking detection or the segmentation task or any object detection task. Rather, you’re doing end-to-end learning, trained on actual behavior of drivers, when they’re driving the car manually.
G: This is hard to do. It’s not supervised learning.
L: Nice that there’s a lot of data; so it’s hard and easy. … Like more data than you need in a sense. It’s an interesting question; in terms of a mount, you have more than you need. But driving is full of edge-cases. So how do you select the data you train on? What’s the cleverest way to select data? That’s the question tesla is probably working on. If you want to create intelligent systems, you have to pick data well.
G: The definition I like of Reinforcement learning versus supervised learning; in the latter, the weights depend on the data. In Reinforcement learning, the data depends on the weights. … We’re not there yet, but that’s the eventual … … .We’re going to do RL on the world. Every time the driver makes a mistake … we’ll train on that.
L: For now you’re not doing the Elon style; promising it’s going to be fully autonomous. You’re sticking to Level 2; it’s supposed to be supervised.
G: Definitely. … look at our rate of improvement in terms of disengagement … maybe we’ve seen 10x improvement in a year. A hundred miles is a far cry from the 100,000 you’re going to need. You need to get three more 10x’s in there.
L: Your intuition is that there’s exponential improvement to be found in there …
G: I think these systems are going to exponentially improve. But it’s still going to take a while, because the gap between our best system and humans is still large.
Tesla vs Comma.ai
L: That’s a good distinction to draw … look at the way Tesla’s approaching the problem versus you … the driving task is an ML problem … Tesla makes it a multitask problem. .. you have this multi-headed neural network that’s good at performing each task. There’s presumably something on top stitching things together, making policy decisions about how you move the car. There’s a brilliance to this … allows you to master each task. Lane detection … traffic light detection, … area segmentation. Vehicle, bicycle, pedestrian detection. There’s localization tasks in there. Also, predicting … how the entity’s in the scene are going to move. Everything’s basically a machine learning task … it’s nice because you can have this entire data engine that’s mining for edge cases for each one of these tasks. You can have engineers that are basically masters of the world. .. .You talk about the Cohen guy for Waymo … become the best person in the world at cone detection. …. Automating much of the process of edge case discovery and training the NN for each … perception task. You’re looking at it from a more holistic way; doing end-to-end learning on the driving task … trained on the data of actual driving of people who use comma ai … plus the moments of disengagement that, maybe with some labeling could indicate the failure of the system. .. You have data for the successful control of the vehicle … and then failure cases where the vehicle does something wrong where it needs disengagement. Why do you think you’re right and Tesla is wrong on this? Do you think they’ll come around to your way?
G: IF you were to start a chess engine company, would you hire a bishop guy?
L: This is Monday morning quarterbacking. Yes, probably. (George laughs)
G: Oh, our rook guy; we stole the rook guy from one company.
L: Well, there’s not many pieces, right? There’s a few that work on the bishop …
G: But is that not ludicrous today, to think about? in the world of Alpha Zero?
L: Fundamental question—how hard is driving compared to chess? … End-to-end is the right solution …
Driver monitoring
Communicating uncertainty
Tesla Dojo
Tesla Autopilot big rewrite
How to install the Comma Two
Openpilot is Android & Autopilot is iOS
Waymo
L: What are your thoughts about Waymo, its present and its future? I’ll say something nice; I’ve visited them a few times; I’ve ridden … You kind of imply there’s zero probability they’re going to win …
G: … I think that the product that they’re building doesn’t make sense. You say the Waymo’s are fast. Benchmark it against a competent Uber driver. The Uber driver’s are faster. ..
L: … stuck at a stop sign.
G: I like when my Uber driver doesn’t come to a stop at the stop sign. Let’s say they’re 20% slower than an Uber. You can argue they’re going to be cheaper. Well, there’s already Uber Pool. Those are 15% of rides. Users are not willing to trade off money for speed. So it’s not going to be competitive with traditional ride-sharing networks. Also, whether there’s profit to be made depends entirely on one company having a monopoly. … the scooter market. Who’s doing well in that market? It’s a race to the bottom …
L: It could be closer to an Uber and a Lyft.
G: Well, the scooter people have given up trying to market scooters as a practical means of transportation; they’re just … super-fun to ride. But from an actual transportation perspective; I don’t think scooters, nor Level-4 autonomous vehicles, are viable.
L: If Elon Musk went on vacation for a year, and the board decides they need someone to run the company—how would you change what they’re doing.
G: I wouldn’t change much. Some minor changes. My debate with them, end-to-end versus seg nets. That’s just software, who cares? It’s not like you’re doing anything bad with seg nets; that’ll help you debug the end-to-end system, a lot.
L: I presume you would, in the Model Y or the Model 3, you’d add driver sensing with Infrared.
G: Yes. I would add Infrared … right away, to those cars. I think they’re already kind of doing it. …
L: Why are you so quiet on Twitter. What’s your social network presence like? On instagram, you do live streams. You understand the music of the Internet, but you don’t always engage
G: I used to have a Twitter. Instagram is a pretty place; it glorifies glory. Twitter, glorifies conflict. … Taking shots at people; Twitter and Donald Trump are perfect for each other!
L: So Tesla’s on the right track, in your view. So if you ran Waymo … they seem at the head of the pack of that approach
G: Level 4 robo-taxi, all in before making revenue.
L: They’re at the head of (that) pack. If you could run this company for a year …
G: I would get Anthony Levandowski out of jail, and put him in charge of the company.
L: (laughing) Do you want to destroy the company by doing that? Or you mean you like renegade-style thinking that throws away bureaucracy … what do you mean?
G: He’s a genius; he would come up with a much better idea of what to do with Waymo … Absolutely, without a doubt, he’s a genius. I’m not saying there’s no shortcomings … He’s willing to take—who knows what he’d do with Waymo? He’s far more out there than I am?
L: I was going to talk to him on this podcast; what do you make of him? I see the best in people; I slowly started to realize, there might be some people out there that have multiple faces to the world; they’re deceiving and dishonest. I trust people, and I don’t care if I get hurt by it, but you have to be a little bit careful, platform-wise, and podcast-wise. You think he’s a good person?
G: Oh, I don’t know. I don’t make moral judgements. I mean this about Waymo; I meant that unironically. PRoblem with putting me in charge—Waymo is already $10 billion in the hole … Comma is profitable … I can build a successful electronics company and still not be able to pay back Waymo’s $10 billion;
L: So your thinking is, how can I make them make money?
G: I wouldn’t know how to save it … it’s beyond me … maybe Anthony could do it. I can’t do it. I can talk about how, what I really want to do is apologize for all those corny ad campaigns; here’s the real state of the technology.
L: I have criticisms, though I’m more bullish. IT went into corny mode too early. It’s still a startup; they haven’t delivered on anything. Show off the engineering they pulled off; not do commercials about their little car company. … Do what SpaceX is doing with the rocket launch. Put the nerds, the engineers up front. Show failures, too.
G: I love SpaceX’s … I’m excited to see them succeed. I can’t wait to see Waymo fail. You lie to me, I want you to fail! …
L: That requires the renegade CEO. I’m with you. I still have faith in Waymo for the renegade CEO to step forward.
G: It’s not John Crapcheck; it’s not Chris Everston. They may be good at certain things, but they’re not renegades …
L: Even though we’re talking about billion dollars; they’re still early-stage startups.
G: It’s against everything Sillicon Valley. Where’s your minimum viable product? Where’s your … users? You think you’re too big to fail, already?
L: How do you think autonomous driving will change society? The mission is to solve self-driving. You have a vision of the world, how it will be different?
G: It’s not about autonomous driving, in and of itself. It’s what the technology enables. I like it because it has a clear path to monetary value … As far as that being the thing that changes the world … who’d have thought … REally the product you’re building is not something people couldn’t have imagined. Could people have imagined the Internet 50 years ago? Only visionaries! But autonomous cars?
L: I told you my long-term dream: robots with whom you have deep connections. There’s different trajectories towards that. I’ve been thinking of launching a startup. I see autonomous vehicles as a step toward that. I think Waymo or Tesla could pivot toward .. that broadly …
G: Gotta stay focused on the mission. For however much I think comma is winning … not until we solve Level 5
L: Before Covid, I’d been thinking of moving to San Francisco. And now I’m thinking about potentially Austin. We’re in San Diego now. So …
Autonomous driving and society
Moving
L: Why San Diego? .. In your case, QualComm and Snapdragon, that’s an amazing combination.
G: That wasn’t why. QualComm was an afterthought. You can have a tech compnay here. …
L: Why does SF suck?
G: People like the ocean,. California, for its flaws … a lot of tthem are SF-specific. First-tier cities in general have stopped wanting growth. The voting class always builds to not build more houses. It is insanely corrupt. Not balanced at all, political party wise.
L: For all the discussion of diversity, it lacks that in thought, background, approaches, strategies, ideas. It’s kind of a strange place. The loudest people about diversity, and the biggest lack of diversity.
G: IT’s the projection!
L: It’s interesting. Even high-up Silicon Valley people say, this is a terrible place.
G: Coronvirus really killed it. SF was the number one exodus place during it. We’ll see what happens with California a bit longer term. Austin’s an interesting choice. Except for the extreme heat in the summer. But that’s very on the surface. I personally love Colorado. You have these states that are, I think, way better run! California, especially SF .. it’s on its high horse.
Advice to Startups
L: What’s it take to build a successful start-up?
G: I haven’t done that. Talk to somebody who did that.
L: …. … Another book for $67 … How I jailbroke the iPhone, by George Hotz
G: And you can too!
L: So you have built, a very unique company. Not you, but you and others. You haven’t really sat down and thought about … If you and I, we’re having some beers, you’re seeing that I’m depressed. There’s no advice …
… L: One of the things I’m considering start-up wise … it’s a technology problem, a platform problem. You said most people don’t comment; but I think most people want to comment. But all the assholes are commenting … That’s a platform problem
G: I can’t believe what Reedit’s become
L: Interesting, because there’s subreedits. There’s small ones that are still havens of joy, positivity, nuanced discussion. Small little pockets. It’s emergent … Naturally, something about the internet; if you don’t put in a lot of effort to encourage positive god vibes
G: I’d love to see someone do this well.
L: I feel like Twitter could be overthrown
G: The Actual Bot talked about how … if you have “Like” and “Retweet,” that’s only positive wiring. The only negative possibility is a comment. That assymetry gives it its toxic-ness. YouTube comments … maybe better; they don’t show the down votes.
L: The … explore possibilities, and get a lot of data on it. The point is, it’s unclear; it hasn’t been explored in a really rich way.. It’s a technology problem. We’ll look back at Twitter as it is now; more likely, somebody overthrows them. “I can’t believe we put up with this level of toxicity.”
G: Advertising as a business model. … The Social Dilemma … the product being sold is influence over you! The product being sold is influence on you! That can’t be; if that’s your idea … it can’t not be toxic.
L: Maybe there’s way to spin it, giving more control to the user, of seeing what’s happening to them. …
G: But… nobody’s going to use that. You can’t depend on self-awareness of the users.
L: You can’t depend on it, but you can reward self-awareness. The ones willing to put in the work of self-awareness … incentivize, and be pleasantly surprised how many are willing to be self-aware. … I’m putting in a lot of work, with you, right now. For an introvert, it’s very costly. But on the Internet, fuck it! I don’t care if this hurts somebody …
G: So much of the engagement is so disingenuous. … You’re doing it to straight-up manipulate others …
Programming setup
L: What’s your perfect programming set up?
G: Give me a Macbook Air, sit me in a corner of a hotel room
L: You don’t fetishize; you don’t want multiple monitors
G: They’re nice; but do they unlock produtivity? No …
L: What about IDE? Which operating system? What text editor? Is there something that … the perfect productivity setup
G: It doesn’t matter! I code most of the time in VIM. Literally, an editor from the 70s. VSCode is nice for reading code. You can build much better tools …
L: I still use EMex. I have to confess: I’ve never used VIM. I’m afraid my life has been a waste. I’m not evangelical about EMex.
G: This is how I feel about Tensor Flow versus PyTorch. … I can’t believe how much better PyTorch is ….
L: My heart; I fell in love with Lisp; I don’t know why. The heart wants, what the heart wants. Maybe … so many AI courses … were taught with Lisp in mind. AT the same time, why am I not using a modern ID?
G: They’re not much better
L: At the same time, I like multiple monitors. The limits of a laptop is a pain in the ass. I’m also addicted to Kinesa’s weird keyboard.
G: At work I have three monitors … I have a (special) keyboard)
L: What about a perfect day, productivity-wise. Not Hunter S. Thompson drugs; what’s the day look like, hour by hour. Any regularities that create a magical George Hotz experience?
G: I remember days vividly where I went through radical transformations to the way I think. The days have been so impactful … My mind was blown. Discovering the Hutter Prize; AI was just compression. Discovering AIXI … loss-less compression implies intelligence! The third one is controversial—a blog called Unqualified Reservations, it’s by Curtis Yarvin.
L: I’m supposed to talk to him, people tell me. He sounds insane!
G: OK, the world makes sense now; I had a framework now to interpret this. … Those days when I discovered these new frameworks.
L: So … what was special about those days … how did they come to be? Did you just get lucky? Did you encounter the Hutter Prize on Hacker News or something like that?
G: I don’t think it’s that, I could have gotten lucky at any point, it’s like—
L: You were ready at that moment … But is there anything magic to the day today. It’s the mundane things …
G: I drift through life hoping and praying I’ll get another day like those days
L: There’s nothing in particular you do to be a receptacle for it
G: No … I didn’t do anything … I did a month-long trip to New York … it was pretty terrible. I did a two-week road trip. Gunnarson, Colorado .. the snow starts coming down. I had to turn my car around. I watched an F-150 go off the road. That day was meaningful. That was real! I had to turn my car around. It’s rare that anything real happens in my life.
Ideas that changed my life
L: Something about that moment felt real. To break apart the three moments you mentioned … Aliza Lkowski … how did your worldview change in starting to consider the exponential growth of AGI that he thinks about. What was so magical to you?
G: Today everyone knows him for threats in AI safety. This …. ….. it was this realization that computers double in power every 18 months, and humans do not. They haven’t crossed yet; but …
L: Did that open the door to the exponential thinking. That with technology, we can transform the world.
G: It opened the door to human obsolescence, the realization that in my lifetime, humans are going to be replaced!
L: The Hutter Prize, I’m torn. What I think about it. The basic thesis; we can reduce the task of upgrading an intelligent system into compression. You can think of all intelligence in the world to a kind of compression. Do you still find that a compelling idea?
G: It’s not that useful day-to-day. Before that, I was searching for a definition of the word “intelligence.” But I definitely have a definition of the word “compression.” Loss-less compression. That is equivalent to intelligence! I’m not sure how useful; but I now have a framework …
L: He just 10x’d the prize for that competition. You ever take a crack at it?
G: I spent the next six months of my life trying it! That’s when I started learning everything about AI. … Then “OK” … I’m caught up to modern AI. I had a good framework to put it all in. Some of the first …
L: BTW, on the compression side … what do you make of the loss-less requirement. Human intelligence and NN’s can probably compress stuff pretty well without that …
G: You use an arithmetic encoder to make it loss-less. Even if you can’t predict exactly the next character; the better probability you have for the next character; you code those probabilities …
GPT-3
L: On that topic, what are your thoughts on GPT-3 … as an AI researcher …
G: I think it’s overhyped. But no, we can’t just scale up to GPT-12 and get general purpose intelligence. The loss function is literally just cross-entropy loss on the character. That’s not the loss function of general intelligence …
L: Can you imagine, to play devil’s advocate, that it’s possible GPT-12 can achieve general intelligence with something as dumb as this loss function
G: It depends. Another problem with the GPTs is that they don’t have long-term memory. So, just GPT-12, a scaled up version of GPT-2 or 3; I find it hard to believe.
AGI
Programming languages that everyone should learn
How to learn anything
Book recommendations
Love
Psychedelics
Crazy
#133: Manolis Kellis, a third visit
First visit here. Second visit here
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Molecular basis for human disease
L: So your group at MIT; is trying to understand molecular basis for human disease.
M: It’s the most complex challenge in modern science. Human disease is as complex as the human genome; as complex as the human brain. It is, in many ways, even more complex, because the more we understand disease complexity, the more we understand genome, antigenome, cancer complexity, and so on. Traditionally, human disease was following basic biology. You’d model organisms: mouse, fly, yeast; you’d understand mammalian, animal biology … in progressive layers of complexity, getting closer to human phylogenetically. You’d do perturbation experiments in those species. Based on the knocking-out of those genes … understand the function of those genes … If you find the function of the gene through the model of the organisms …. You would start with mammal models, eukaryotic models. But this has all changed. Human genetics has been so transformed in the last decade or two. It is driving … There is more mutation information … Perturbations is how you understand systems. An engineer builds systems and knows how they work from inside out. … If I polk that balloon, what’s going to happen. You can make expereimentation by perturbation. The scientific process is building models that best fit the data … It’s the same with science; if you’re trying to understand biological science …
L: How do these perturbations allow you to understand disease?
M: You want to know, what is the disease mechanism; then you go and intervene. Traditionally, epidemiology, the study of disease, the observational study … has been about correlating one thing with another. If you have a lot of people who have liver disease who are also alcoholics … hypothesis. … With genetic epidemiology, it’s about correlating changes in genome with phenotypic differences; then you know the casuality (directional). . .. Okay, perturbing that gene in mouse causes the mouse to have X phenotype. … Perturbing that gene in human .. causes disease. … You can prescribe medications that alter these processes
L: That’s such a beautiful puzzle to try to solve. You study that for mice first; and see how that might possibly connect to humans. How hard is that puzzle?
M: In animals, we make the puzzle simpler, because we perturb one gene at a time. That’s the power of animal models. We can de-couple the perturbations. We do one strong perturbation at a time. In humans, the puzzle is incredibly complex. You don’t do human experimentation. You wait for natural selection to do its own experiments in the human population. It’s been doing it for hundreds of thousands of years. You take this natural genetic variation that we all carry within us. Every one of us carries six million perturbations. … I’ve done six million experiments on you, on everyone. …
L: Six million?
M: Six million unique genetic variants that are segregating the human population. Every one of us carries millions of polymorphic (means many forms or variants) sites. These can be thought of as tiny little perturbations. Most of them don’t do anything. Some of them lead to all the phenotypic differences we see between us. The reason two twins are identical
L: How happy are you with this kind of data set? Is the human population of earth …
M: That’s a power analysis question. … What is the effect size I’d like to detect? What is the natural variation between the two forms. … Form A has some natural phenotypical variation around it. Same with Form B. … The further the means go apart, the more power you have … Is that sufficiently large? Certainly not for everything? But we have enough power for some of the stronger effects …
L:
M: The allelic ___ of a.gene …
L: You said in the past, there would be small experiments … on animal models. What does this puzzle-solving process look like today
M: We have seven billion people on the planet. That’s an enormous matrix of genotype, by phenotype, by systematically measuring the phenotype of individuals. The traditional way has been by looking at one trait at a time. Gather families … The pedigrees of a strong effect … Mendelian … Where basically one locus plays a very big role in that disease. Look at carriers versus non-carriers in one family; in another family; in hundreds, maybe thousands of families.
L: Is this the matrix you’re showing in talks or lectures?
M: That matrix is the input. That’s traditionally been strong-effect genes. Now, instead of families, or pedigrees, you basically have thousands, or sometimes hundreds of thousands of unrelated individauls—each with their phenotype, for example, height, or lipids, or sick or not. That has been …
M: … we’re doing this in the context of large medical systems are very-large cohorts … sometimes with their complete electronic health record. Sometimes you can be relating one gene … one family; Now you can do millions
M: … for one genetic variant to affect Alzheimer’s; that’s a very long path. These six million variants that each of us carries; that one single nucleotide … has this effect … all the way to the end.
L: So, the most effective place to detect the alteration that results in disease is as early as possible in the pipeline.
M: It’s a tradeoff. Early on, each epigenomic alteration … is active 50% less. Now you can ask, how much does one change … change disease; that path is now long. .. If you look at expression, it’s a subtler effect at the gene level. Now you’re closer; you’re looking at one gene in the context of 20,000 other genes; as opposed to one enhancer in the context of two million other enhancer. ….. … What is the … influx rate when I have this genetic variation? What is the synaptic density? What is the electric impulse conductivity? You can measure things along this path to disease. You can also measure endo-phenotypes. You can measure your brain activity; imaging in the brain. The heart rate, pulse, lipids, amount of blood secreted. Through that, you can look at the path to causality, to disease.
L: Is there anything beyond cellular? You mentioned lifestyle interventions or changes, being able to prescribe that. What about organs? What about the function of the body as a whole?
M: When you go to your doctor, they always measure your height, weight, BMI, etc. With digital devices nowadays, you can measure hundreds of things for every individual … You can basically also phenotype cognitively through tests, Alzheimer’s patients. … There are phenotypes that we do for cognitive decline. … Enlarging the set of cognitive tests. … How do they get out of mazes … In the human, they can have much much richer phenotypes. .. Not just imaging at the organ level; all kinds of different activities at the organism level. … You can do behavioral tests; how did they do on memory? On empathy?
L: I love how you’re calling that phenotype … behavior patterns that might change over the duration of a life. Your ability to remember things; your intelligence perhaps.
M: Yeah but intelligence has hundreds of variables!
L: We’re able to measure that better and better
M: We used to think of it as a single variable. That’s ridiculous! It’s dozens of different genes that are controlling every single variable. Imagine us in a videogame; every one of us has measures of strength, stamina, energy-left. … Each bar will give you hundreds of bars. For my Machine Learning task, I want a human who has these particular forms of intelligence. I require now these 20 different things. Combine these things, and relate them to performance of a task; but also to genetic variation that might be affecting different parts of the brain. Genetic variation that affects expression of genes … affects music ability, smell, dozens of different phenotypes, that can be broken down into hundreds of cognitive variables, each of which is related to thousands of genes …
L: So someone who loves RPGs, there’s too few variables that we can control. As a game player, I’m excited by the quality of the video game. The designer did good.
M: I don’t know; the sunset last night was a bit unrealistic …
Deadliest diseases
L: What are the most important diseases to understand?
M: There’s many metrics of importance. One is lifestyle impact. If you look at Covid, its impact has been tremendous; so it’s very important. All of these impacts for millions of Americans. That’s one aspect of importance. Mental disorders—Alzheimers’ has a huge impact, whether or not it kills someone. So the first measure of importance is just well-being. The second, easiser to quantify, is deaths.
L: What is the number one killer?
M: Heart disease. 650,000 Americans per year. Second is cancer, at 600,000. Third is accidents—every accident combined. Car crashes. 167,000. Lower respiratory disease—asthma … 160,000. Alzheimer’s, 120,000. Strokes/aneurysms. 107,000. Diabetes, 85,000. The flu, 60,000. Suicide, 50,000. Overdose, etc. Of course Covid has creeped up to be No. 3 this year, with more than 100,000 and counting. But if you think about, what are the most important diseases, you have to understand (both measures).
L: EAch of these diseases you can think of, along with terrorist attacks, as things that lead to fatalities. You can look at as problems that need to be solved. Some are harder to solve than others. If you look at these diseases, heart disease, cancer, Alzheimer’s … schizophrenia, obesity … things that affect quality of life? Which are solvable? Which aren’t?
M: I love how your question puts it in the context of a global effort. From the global aspect, exercise and nutrition are things we as a society can make a much better job of. The availability of cheap food … if we change that equation; and as a society, we make availability of healthy food much easier, and charged a burger at McDonald’s—the price it weighs on the health system, people might get healthier food. In the same way, increasing empathy or education … would lead to fewer suicides and murders. That’s something we as a society can do. You can think about external vs. internal factors. .. Communicable diseases are external, where as cancer is internal—your genetics will eventually drive you there. For each, there are genetic and environmental components. … Alzheimer’s … 79% heritability… that’s a big component. Yes, there’s a 21% environmental component … you could enrich that environment, read more books, go running, have a more fulfilling life; that will decrease Alzheimer’s, but there’s a limit to how much that can impact.
Genetic component of diseases
L: So each of these problems have a genetic and environmental component. When it’s genetic, what can we do about this. What is solvable?
M: So my group works on the genetic component, but I would argue … if we couple the genetics with expression variation, and ask—what genes change by a lot; which gene correlate with disease by a lot … those are the best places to intervene. Pharmaceutically, even a modest effect will have a huge effect on the disease.
L: … What have we been able to find, in terms of which disease can be helped?
M: Don’t get me started! Our understanding of disease has changed dramatically. One of the worst things about my genome, is I have a predisposition to Age-Related Micro Degeneration (AMD). A form of blindness that causes you to lose … vision. My increased risk is 8%; you have a 6% chance.
L: You know this about you?
M: I know this about me.
L: Philosophically, that’s a powerful thing to live with … We’re going to try to focus on science today and more philosophy next time, but it’s interesting … the more you’re able to know about yourself from genetic information, how that changes your own view of life.
M: There’s something called genetic exceptionalism that thinks about genetics as very different than everything else. And, you know, let’s talk about that next time. Going back to AMD, we have no idea what causes AMD. It was a mystery—until, the genetics worked out. Now that I know that I have a predisposition, it allows me to make some decisions. …
Genetic understanding of disease
M: … schizophrenia …
L: … the synaptic pruning process … How did you figure this out? This intricate connection, pipeline of connections
M: Another example: Alzheimer’s disease. You’d expect it to act first in the brain. We had this roadmap epigenomic consortium view of the epigenome … across 127 different tissues and samples, with dozens of epigenomic marks over hundreds of doners. We learned, you can map the active gene regulatory elements for every one of the tissues in the body. We mapped … for what regions of the genome are turning on … in different tissues. We can ask, where are the genetic loci that are associated with disease. This is something my group was the first to do; we were able to show that specific epigenomic states were enriched … we pushed that further in the Nature paper a year later. In this roadmap paper a year after that. That matrix that you mentioned earlier was the first time we could see what … traits .. .. .. A lot of that map made complete sense. If you look at … allergies, Type 1 Diabetes; the genetic … were enriched. … Enhancers .. active in T-cells and B-cells. That gave us confirmation, in many ways … were enriching immune cells. If you look at Type 2 Diabetes .. there was an enrichment only in pancreatic eyelets. … regulation of insulin … in pancreatic eyelets. That was spot-on super-precise. If you look at blood pressure, where would you expect it to occur? .. We found it localized specifically in the left ventricle of the heart. The enhancers there contain a lot of genetic variances … … Height … is acting in developmental stem cells; makes complete sense. If you look at inflammatory bowel disease; that’s both immune and digestive. We saw double enrichment in both immune cells and the digestive cells. So both components exist … Big surprise in Alzheimer’s. We found zero enrichment in the brain samples … What is going on? What’s going on is that the brain samples are primarily neurons: neurocytes and astrocytes, in terms of the cell types that make them up. So … genetic enhancements associated with Alzheimer’s … was acting up only in the fourth type of cell: micro-glea
L: Oh, nice
M: (Laughs) … sell-surface markers of those cells. Just like microphages that are circulating in your blood, the microglea … are tissue-specific monocytes. Every one of your tissues … …. has a huge role to play in obesity. Again, these immune cells are everywhere. So, with this completely unbiased view, we found that Alzheimer’s was tremendously enriched in microglea but not at all in the other cell types.
L: … is that simply useful for … does it give us a pathway of treatment?
M: Very much the latter. You have to start somewhere. You are going to make asses that manipulate those genes and those pathways in those cell types. Before we know the tissue of action, we don’t even know where to start. But if you know the tissue, and even better the pathway, of action. Then you can develop a … robotic system for testing the impact of your favorite molecules, that you know are safe, efficacious; you can screen them against, a set of genes that act in that pathway, or on the pathway directly. You then go into mice, do experiments; and figure ways to … then go back to humans, do clinical trial, try to do the same thing in humans. But that’s not the only part. Many additional steps to figuring out the pathway of disease.
L:
M: I mentioned a lot of diseases … If you look at biology, it used to be zoology and botanology and virology departments. MIT was one of the first to create a biology department—we’re going to study all of life, suddenly. … The central dogma of DNA makes RNA makes protein—in many ways unified biology. …
Unified theory of human disease
M: It’s inevitable. Fundamental understanding of circuitry of human genome that you need to solve schizophrenia; but that same circuitry is hugely important for Alzheimer’s, and immune disorders, and cancer, and every single disease. All of them have the same sub-task. I teach dynamic programming in my class. Dynamic programming is about not re-doing the work; it’s about re-using the work that you do once. The idea of having different specialists solving the same problem separately, it’s crazy. What we need is a circuitry group to come up with the circuitry of everything, applying it to all disorders associated with immune dysfunction. All of them will be interacting with the circuitry folks. That’s the current structure of my group, if you wish. We’re focusing on the fundamental circuitry. … Collaborating with many other labs. We have a hard focus on cardiovascular disease; we have an immune focus on several immune disorders. Cancer focus on metastatic melanoma, and chemotherapy response … psychiatric disorders … Alzheimer’s, ALS, AD-related disorder, dementia, … a metabolic focus on the role of exercise and diet and how they are impacting … organs across the body. All of them are interfacing with the circuitry, and … another CS principle: eat your own dog food! If everybody ate their own dog food, dog food would taste a lot better. The reason why Power Point and Microsoft Office was so successful, was that the employees working on them were using them for their day-to-day task. You can’t produce the circuitry if you’re not simultaneously users of that circuitry … We are building the map of what are the genes in the human genome. Over the years we’ve written a series of papers on how do you find human genes in the first place? How do you find how these motifs come together? How do you link regulators to enhancers and enhancers to their target genes? Using epigenomics! We’ve built all this infrastructure for understanding every single nucleotide of the human genome, and how it acts in every tissue of the human body. That’s something my group has taken on; no small task. We have also worked with specialists in every one of these disorders; in some cases collaborating with pharma to go all the way down to therapeutics. … that basic circuitry, and how it allows us to improve the circuitry; not just treat it as a black box. We’re focusing on that because we’re understanding … from the disease front.
Genome circuitry
L: Maybe … one nice question to ask ; how do you go from knowing nothing about the disease, to going through the pipeline and actually having a drug or treatment that cures that disease.
M: That’s an enormously long path and great challenge. I’m trying to argue that it progresses in stages of understanding, rather than one gene at a time. It used to be that one post-doc would be working on one gene, and that’s their job. We’re now seeing how poly-genic the diseases are. So we need to have cross-cutting … needs. Every single one of these paths we’re now doing in parallel across thousands of genes. First, we have ………… gene regions, and ultimately genes that are underlying particular disorders. The polygenic path … looking at individuals … finding hits, where a particular variant impacts disease all the way to the end. .. Now we have a connection, between a genetic region and a disease. Not the same as having a connection between a gene and a disease! The reason for that is that 93% of genetic variants that are associated with disease don’t impact the protein at all! There’s 20,000 genes and 3.2 billion nucleotides in the human genome. Only 1.5% of that encodes for proteins. If you now look at where are the disease variants located? Ninety-three percent of them fall outside the … proteins. The problem is that when a variant falls outside the gene, you don’t know what gene is impacted by that variant. Because the genome circuitry is very often long-range. So .. that genetic variant could sit in the “entron” of one gene. Genes are split up into ex-ons and entr-ons. … It’s transcribed with a gene; but it might not impact this gene at all.
L: So it’s just riding along …
M: An example. Genetic association with obesity was discovered in this FTO gene; which was then studied ad nauseum. They figured out that FTO was RNA-manipulation … just like the genome can be modified, the transcripts of the gene can be modified. We thought, oh great! … My group studied FTO in collaboration with a wonderful team led by Melina … This FTO locus, even though it is associated with obesity, does not implicate the FTO gene. … It controls two genes; IRX-3 and IRX-5 that are sitting 1.2 million nucleotides away.
L: What am I supposed to feel about that?
M: Well, when I … I wrote the most depressing paper of 2015. People had become comfortable that there was just one single gene in that locus. … Our paper showed that you can’t ignore the circuitry. You have to figure out the circuitry, all those long-range interactions … And then, you now have one of the building blocks, for going in, and understanding disease.
L: So, embrace the wholeness of the circuitry. Back to the question of knowing nothing of disease and going to the treatment …
M: First figure out the tissue ….. … figure out the epigenomic enrichments. The reason for that, thankfully, is that there is convergence. The same processes are impacted, different ways by different loci. A saving grace for our field …
L: Can you clarify why that’s hopeful
M: There’s a small number of biological processes that play the biggest role in every disorder. With Alzheimer’s, one of them is lipid metabolism … These are a small number of processes; but with multiple genetic perturbations associated with them. … synaptic pruning … When you look at disease genetics; you have hits in various places … it turns out these hits are calcium signaling proteins … Focusing on one gene at a time, you can’t see that picture. You have to look at the holistic picture …
L: You mentioned the convergence thing …
M: Convergence is such a beautiful concept! You basically have these four genes that are converging on calcium signaling. Each acting in their own way, but together on the same process … But at these loci, you have various enhancers controlling each of these genes. In each of these enhancers, you might have multiple genetic variants, distributed over many different people … schizophrenia …
L: Saving grace; that happens in a lot of these diseases
M; For all of them. .. some epigenetic enrichment … For 540 disorders, we found that there is indeed an enrichment; there is commonality, and from the commonality, we can get insights. We’re basically building an empirical prior. A Bayes-ian approach … In a genetic locus, you have a dozen …
CRISPR
M: CRISPR is this genome-guiding and cutting mechanism. You can basically guide RNA that you put into the system … find wherever there’s a match, and cut the genome. Bacteria are constantly attacked by virus. When they win against the viruses, they chop them up; they have this CRISPR loci … stands for Cluster Repeats Interspersed etc. … segments that were basically matching viruses. This is probably a bacteria immune system that remembers the trophies of the viruses it manages to kill. It passes on these memories; the CRISPR system was adapted as an immune response to sense foreigners that should not belong, and cut their genome. It’s an RNA-guided, DNA-cutting enzyme. Different systems; some cut DNA, some cut RNA; all of them remember this sort of viral attack. What we’ve done now, as a field, is co-opted that system of bacterial immune defense as a way to cut genomes. You have this guiding system … an RNA guide … to cut DNA at a particular focus.
L: Wow! So we recognized, we can use that think, it’s a nice tool.
M: IT was discovered by the yogurt industry; trying to make better yogurt, and trying to make bacterial culture in their yogurt more resistant to viruses … our DNA has a repair mechanism …
Mitochondria
L: What are the technologies that made this possible? What are we excited about?
M: I love that question; there’s so much ahead of us. So solving that one locus required massive amounts of knowledge that we’ve been building … to figure out causal variant. It required knowing this regulatory genomic wiring … It required … genetic perturbation of these gene phenotypes. All this put together for one locus. A massive team effort; huge investment in everything. Just one paper … This one single locus. I like to say it’s a paper about one nucleotide in the human genome. .. We have to go through 3+ billion nucleotides. How to do it systematically? The technology actually allows us to do it thousands of loci at a time. What technology allows this? What is automation and robotics. You can screen through millions of molecules, thousands of plates, each containing thousands of molecules, every single time testing all of these genes. Asking which molecules perturb these genes? … micro-array technology … measure gene expression by hybridization …
Future of biology research
The genetic circuitry of disease
…
M: .. computer science comes together with … technologies .. that are revolutionizing the way of intervention … massive … There’s no better way, there’s no better time/place, to be looking at this whole confluence of ideas; I’m just so thrilled to be part of this amazing, enormous ecosystem.
L: It’s exciting to consider the life for humans 100-200 years from now. .. When they look back at us; they probably wonder how we put up with all this suffering in the world. Manolis, it’s a huge honor.
#134: Eric Weinstein, pre-election visit
He is the wise turtle to my Kung Fu Panda … His podcast, The Portal. I’m drawn to the possibility of having many more conversations with Eric; we have the right kind of contrasting worldviews; and a deep respect for each other’s life stories. This magical experience in the realm of conversation; we’re always looking for something that we never quite find, but we’re always better for having tried. I’m not sure how the world connected us …
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Eddie Van Halen
L: Who’s the greatest musician of all time? We were just off-camera talking about Eddie Van Halen; he just passed away.
E: Jonathan Richmond. A front man for a group called “The Modern Lovers.” We have to be prepared to play music, when the electricity’s out … I thought it’s … We were having a conversation now about virtuosity … that (Van Halen’s death) affected me emotionally. I revered EVH’s capacity for innovation. What it’s like to totally discontinuously innovate. The arpeggios that he did on a single screen; so fast. The attacks for the hammer-ons …
L: You’re often so poetic about music; it clearly touches your soul on many levels? … Is that deeper than just rocking out in your convertible? I imagine EW just driving down California highways blasting some kind of music. Is it just being able to be carefree for moments of time? Or is it something more fundamental?
E: How often do you have the chance to have math performed, as you do in Bach? … To go back to Leonard Cohen’s famous line, “The Baffled King composing.” Individual verses of that song are insanely important. The Baffled King is how we often make music. We don’t understand, what did we just do to break that person’s heart sitting on the couch? You’ve got this weird open-music port on the computer! Somebody’s playing guitar and they’re making you feel things. Particular instruments, like the violin. It’s so difficult, so unforgiving; when it gives up its secrets. .. Sometimes I talk about “head, heart, and loins.” When something can grab those things at the same moment; there’s very few opportunities to live like that. Your head, the musical innovations; he was drawing directly from the classical cannon. Maybe rock is what somebody like Jimi Hendrix saw it as being. In terms of heart, I always notice the smile on his face. It’s painful to look at an EVH solo now; seeing the cigarette dripping off the side of his mouth. I like knowing that (he’s) in the world; I’ve never heard a guitarist say, “Eh, I think he’s okay.” You can hate him; but you still think he was a genius. And then loins; those leaps? That guy was incredibly good-looking. He completely owned the male sexuality of the stage. Both being the completely dominant mythical alpha male; I hate that expression, but there you are. The sense that it all came together? How could you not eat that up?
L: You can imagine the million teenage boys, playing air guitar in their room, dreaming of being that kind of god; the most perfect example of what a human being can be.
E: As in many of the cases with these bands; the original configuration with David Lee Roth. DLR is such a hot mess at all times.
Leonard Cohen
Battle between ego and humility
Darkness and beauty
E: … once you realize there’s perfection and an inability to make contact with perfection, you recognize that there is no solution to this world.
L: That’s weird, with the poets and musicians. You want to say, this is a particular thing that you do. Then there’s Spanish Fly by Van Halen.
E: … I think SF is very singular
L: I couldn’t imagine Eddie Van Halen separate from the band … it made me imagine him sitting alone on a couch in a room
E: I think that’s who he was! I really do! I get it; he was a rock star. I’m almost positive you can’t get to where he got to, without being a complete introvert.
L: It made me imagine there’s a half-naked supermodel, walking around, hoping they could do their thing together; and he’s completely disinterested
E: … the guitar. … maybe you’re conquesting; you’re pursuing love and romance. OTOH, you’re talking about a relationship to the Creator, to … reality. Do I beleive that Eddie Van Halen and Jimi Hendrix and … jacked in to the true essence of the world? Yeah! I don’t think it’s as good as differential geometry. I think it’s .. beautiful, for other reasons. Thank god, because it’s difficult to communicate differential geometry at scale. … You could put him on wires, and set his pants on fire, whatever. .. OTOH, you want something completely precise that shows off … what’s possible with the Stratocaster
L: There’s a precision to it which is very different than Hendrix …
Jimi Hendrix
L: There’s a messiness to Hendrix …
E: … maybe every generation has to have its Hendrix conversation ….
L: There’s so many details. One: It hurt my soul on so many levels that you can put a thumb over the guitar to hold a note; it doesn’t, because I wanted to be the Russian virtuoso that sits with his guitar in perfect form. You want the thumb to be perfectly relaxed
E: That’s the Russian conservatory student. Then there’s the Russian wild man. There are different Russian archetypes. … I can do this backwards, in my sleep, in any time signature … We’ve discussed my piano tuner, in previous episodes.
L: No, that was offline. …
E: There I was, in darkest Manhattan; with a Spinnet piano. It fell out of tune, and I would have to tune it. The only tuner I knew was this Russian guy; his attitude rubbed me the wrong way. He said, “Are you sure you want to tune this piece of shit.” The phone rings, and I have the phone ringer set to Paganini … immediately he figures out what key the phone-ringer is in. It’s not the key for the original composition. He starts going into variations on the song that I’ve never heard before. The phone stops ringing; we have this awkward silence. He then says: “No, you are the piano player; I am merely the piano tuner.” Oh man! Through the heart!
Good Will Hunting
L: It’s reminiscent of the Good Will Hunting story
E: I think about Matt Damon, as a young guy, risking everything, giving up Harvard. I think the most accomplished people in the world are those who choose to give up Harvard voluntarily. Ives was one of those people; Bill Gates of course. And then oddly
L: Zuckerberg
E: But then Steve Jobs gave up Reed. It’s the craziest college in the world. What it was before the current craziness … I think a lot of my reaction is to the real story of Matt Damon, having this vision, and being the young guy to pull it off. Also, Robin Williams trying to explore heart through this lens of acting. … Comedians know they’re a screwed-up bunch of people. They’re proud about it; they really are. The idea that Robin Williams, who I saw in the comedy clubs; he was a straight-up, crazy, disregulated genius. In tremendous pain; his ideal to do it … through acting. Not just being a clown … I was really moved by that. He brought some authenticity; it took a huge risk, for a comedian, to be that real.
L: The madness and the genius were neighbors.
E: That one couldn’t be any other way. Seeing him in a comedy club; he would react to random stimulus in the environment. You almost got the feeling, he wanted a heckler; he was instantly inventive.
L; To me, the best Robin Williams as he got closer to the end of his life; because there was a sadness, and he’s almost fighting the sadness, with this improvisational … the weapons he has is this wit and humor, and this dancing that he does with language. When he falls silence, you can see the sadness; like a bird with a broken wing that’s trying to fly. He would have made one hell of a podcast guest
E: What we call podcasting is just getting to know a soul over and over again. Maybe the idea, talking about depression, and sadness, and heavy feelings, is not an American specialty. Seeing that in context with the beauty of wife—is a Russian specialty.
L: Sounds like a diner menu
E; A scoop of ice cream with tons of depression! I do think we’re in a terrifying and depressing time. We don’t even know what this is! World … falling into some new state; and we’re not even super-curious as to what the hell just happened. Everybody’s got an answer … I’m convinced most of those answers are wrong.
Revolutionary minds
E: The central core of the answer seemed to be that the U.S. was the greatest in the world because we hadn’t noticed that we were getting a benefit from not having to make a plan for low growth. As long as we had growth, we were in great shape. Let’s imagine you could run an experiment; you had a billion copies of earth; you start the initial conditions slightly different. On many planets, the things that were discovered from 1801-1999 were discovered in a period of time—once you crack the puzzle of getting better instruments, you can see more. Turns out there was lots of stuff to do, with germs, or electron orbitals, or . … we got to do a lot of those things. The U.S. roughly corresponded with that bonanza! Of course we look like this genius country. Imagine a car where you don’t have to put in mirrors, seat belts … all of the stuff to keep things from going wrong. I think that’s what we had; as long as growth was insanely good, we plowed riches back into the system, and made more genius stuff; we carried along a good chunk of humanity. We did not have a plan of when the growth goes below the crawl-speed of society
L: How sure are we the growth has slowed?
E: The right concept is … I use this analogy of the orchard, because people talk about low-hanging fruit. So people say things like, “You think we’ve picked all the low-hanging fruit, but I believe in the inventiveness of the human mind.” What if we need to go look for new orchards, or plant new orchards. My claim is, there’s probably lots of low-hanging fruit, but not here. … Some of it turned out to be in the digital orchard. The chemical orchard has gone sort of stagnant
L: I have a faith, that there’s a small percentage of the population that’s looking for those orchards. I’m excited about one of them—there will be robots in everybody’s home, which will unlock some totally new thing. … Everything; I’m excited about that orchard; I’m roaming that orchard, wondering how the hell .. bring back the ant that finds a new source of food. I’m trying to bring an apple …
E: Great … I don’t think there are very many of us. Those that explore
L: it takes one ant! How many does it take?
E: Imagine some ant finds a new source of food; it comes back to the colony. Initial reaction is, you’re not authorized to find new food!
L: WE’re going to remove you from Twitter
E: And by the way, that shows how privileged you are as an ant. .. not a good model for finding new orchards. Where there’s a system that allows somebody to ascend without a lot of gatekeeping, you can have that. … I saw that in hedge funds; hedge funds for a while hoovered up a lot of talent, because there you had funding and you had freedom. REally smart people want to be free, and don’t want to think about how to feed themselves. You can give them productive places to play, … dangerous places … We’re not providing for the people who have to disrupt, who have to innovate. We’re so focused on this other thing: fairness and safety! The singular focus on fairness and safety, while in the same breath being focused on growth—is going to doom us. We’re always talking about divying up the pie that is, as opposed to the pie that will be. .. Imagine forever dividing up the 13th-century pie—you’d be an idiot
L: One place I think I disagree; I don’t think you need that many people to empower the geniuses, the innovators … the people who refuse to spend most of their days in meetings about fairness … I think podcasting, whatever you call that medium, is one little example of a tool you can give power … you on your podcast can have the next Elon Musk
E: There has been a series of places for people to play and be free. And we’ve lost them, successively. … You’ve interviewed Noam Chomsky. He comes from an era where you could play at MIT.
L: I disagree with you! I wasn’t brave enough at that time; I’m not brave enough now. .. People are going to tear me apart … The feeling I have
E: Whose podcast is this—it’s yours? Tell the people who are currently editing your brain that they should go find another podcast. Let’s get rid of some of those people right now. …
L: Nevertheless all the self-doubt is sitting there right now … The thing is, when I walk the halls of MIT; there’s bureaucracy; there’s administrators who never have done anything interesting in their entire lives. .. But in the eyes of individuals, there’s glow of excitement, has nothing to do with career. It’s still a playground! From which geniuses can emerge; they’re unaffected by diversity meetings
E: I love to hear it; I don’t believe it. ‘Cause I’ve watched the change; we’re all editing ourselves all the time. All this relentless focus on CRT, fairness, social justice, it’s making many of us into worse people
L: Do you think the Matt Damon characters are paying attention to any of that?
E: Have you see what happened to Matt Damon himself? He can’t speak.
L: Well let’s not mix up; he’s just an actor
E: No, he’s a Harvard student. He came up with a screen play?
L: You don’t think you can build a rocket company at MIT still?
E: I do think; I think there are things you can still do
L: But we’re losing them
E: We lose them
L: I would say the biggest problem; I think the solution would be, to fire anyone who isn’t faculty; young faculty should have way more power, and administration should have less power. Because right now, the administration has lost the fire, the spark, the memories of the playground. And so the people who admire and love the playground, should have way more power. Create systems that give them power
E: Your’e very idealistic; you’ve got a huge heart. I don’t want to dissuade you from believing beautiful things. I see how potent you are; you do all these things: jiu jitsu, guitar, podcasting, programming computers, etc. I don’t think you’re right. I think we’re in a deeply screwed-up place.
Next Elon Musk should come from MIT
E: I do think there are still people who are capable … that progress the story forward. Look at the fire that some of the people are in who fit that profile. Like how much crap has Elon Musk taken? Quite considerable.
L: And not much admiration, from the
E: Craig Venter. Jim Watson (Editor’s Note: not sure who he meant: this guy?). These are very difficult people! Steve jobs is a very difficult guy.
L: It is a bit heartbreaking to me; my mind is a little focused on Elon Musk because he’s the modern
E: You know him; he’s a person to you
L: It hurts my heart to see how few faculty and people with Nobel Prizes admire Elon. He gets a lot of fans from people who buy his products; young minds, but why doesn’t MIT say that we … somebody amongst us will be the next Elon Musk and we want to encourage them? Like, say that! That’s success! … Instead there’s this jealousy. “Did you hear what Elon Musk tweeted? How irresponsible is what he’s doing?” All these things that are dripping with jealousy
E: I want what he’s got! … Rivalry has a different signature. When you know that you’re never gonna make it, that’s the position you take. What is it, in Kung Fu Panda, that you’ve watched now; what does Ty Lung say when he’s looking for the dragon warrior, and the Furious Five come to defeat him … one of them gives up Po’s name accidentally. “So that is his name; finally—a worthy opponent! Our battle will be legendary!” He’s excited. Why is that? You learn about this in boxing; sometimes you’ll see a division which is lousy with talent; sometimes you’ll have a division which (OTOH) has one star and no real competition in that weight class. That person is in bad shape, because you can’t build a legend without the other. When you think of Muhammad Ali, what names do you immediately think of.
L: Frazier; you have to think of the other heavyweights
E: Those opponents are in part what made Ali Ali. That’s why the Mayweather McGregor revelation; this guy’s got his opponent’s picture in his house. How weird is that? Because without the opponent, you might not be able to get there. .. I’m not a fan of the wrong kinds of rivalries—where people take each other’s credit, and screw each other over. … Examples, the RNA-tie club … (editor’s note; the RNA-Tie club is a group of scientists that included this James Watson) where these guys were so in love with what they were doing, that they couldn’t wait to share everything. .. most people got Nobel Prizes for being a member of the club and doing cool stuff. That’s the golden kind of sweet spot. Most of these people can’t do what Elon’s doing, because they can’t break rules; they can’t take the pressure.
Love will save the world
E: What concerns me about your perspective: I think there’s a lot of genius ideas inside of people who don’t have the stomach for conflict and derision, and I think a lot of those people are female. And I think that until we come up with a world that can swat down the trolls; where we can actually cause the trolls to not ruin everything. I don’t necessarily mean by shutting them up; but somehow separating off people who are working and who are trolling. We are losing a huge amount of genius, in part because women in particular aren’t going to push an idea if it results in 10 years of being derided. Very few men will do that either; but some of us will stick to an idea for 10 years even if the world collapses. I don’t think there are as many women that are going to make that calculation
L: I think technology can help us fight the trolls. I believe a better Twitter can be built
E: I do not (agree). I think you can always improve; but ultimately I think the problem isn’t Twitter; the problem is us. I recently realized … academics and trolls have very many similar behaviors. It’s largely a trolling community
L: I tend to believe the trolls — it’s like the Peter Thiel mini-mind idea. In all of the trolls, there’s the possibility of goodness. What you have to do is create technology that incentivizes them to embrace the better angels of their nature. I believe that, like people actually want to do that. Trolls is a short-term dopamine rush, of childish toxicity that all of us want to overcome.
E: I try to keep myself from believing what you believe.
L; Because you’ll be disappointed
E: It’s dangerous; a lot of these people are implacable foes. You meet them, they’re like, I’m here for the pain.
L: Even in them, there’s a good—
E: Wonderful book I’m going to recommend to you, called Maximum city, about Bombay. The conceit is the author leaves Bombay as a kid, and comes back as an adult … he can’t live in the city he left. He gets in contact with the weird areas of the city; including the underworld. There he’s talking to contract killer. He says, it’s really weird, everyone pleads for their life right before I kill them. They always say, “I’ve got two kids at home.” Never say that to us, because we all have bad relationships with our parents. … You don’t know how people are wired. As much as I hate to say it, there are people whose wiring is so disturbing and different from yours, that you will never guess why you can’t reach them; they might have gone over a point of no return
L; Nevertheless … you’re making a hypothesis. You don’t know this is for sure. I’m whatever the hell I am, with a different hypothesis, that even in the darkest human beings, that seem to be only full of evil, there’s a good person … to be discovered.
E: I love doing your show, because you have these beliefs, even as a Russian. The Russian is total cynicism, and total idealism, right together
Are we headed toward a Civil War?
L: I’m really worried about the next couple of months. If there’s anybody in this world that could help alleviate my worrying, by at least walking along through this worry of mine, it’s you. Do you think we’re headed toward some kind of Civil War; some kind of division that goes beyond stuff on Twitter? Something really destructive..
E: I believe we’re in a revolution … I think waiting for this to be called a civil war is not smart. I think the problem is encountering things you’ve never seen .. and calling it things you already know …
L: But history repeats itself
E: But I don’t see that. Famous quote is that it rhymes
L: I guess .. violence
E: We’re in there; the abstraction of violence. Imagine you were coding up violence as an abstract path. …
L; Thanks for speaking to the audience
E: … You have a smart audience … They’ll digest it for each other; one of the great lessons of long form podcasting. They’re happy doing the work. The people who don’t want to struggle, will leave. I think the point is, you would want to say, violence is defined relative to a context; let’s call it meta-violence. We already have a term for physical violence. I would say physical violence is sub-classed for meta-violence. Meta-violence is the disruption of a system. If a cell dies, it can die through apoptosis or necrosis; the former is controlled cell death. The latter was a violent disruption of the system.
L: This violent class; is it all negative?
E: No! What are you talking about? That’s the madness of our age. If you open up a drawer in our cabinet, and you see knives, forks, and spoons; do you think the spoons are good and the knives and forks are bad? The knife is there to do violence! That’s violence you want done! When you cut a mango, you’re doing violence to the mango. The mango expects I’ll do violence to it. Violence is absolutely part of our story
L: So there’s metaviolence
E: It’s a multiple inheritance pattern …
L: But there’s certain subclasses that allow evil to emerge. What I’m specifically worried about
E: Yeah what’s on your mind, Lex?
L: I worry that amidst the chaos … we have these protests, or the chaos that could be created by the feeling the election does not represent the voice of the people … whoever wins based on some kind of reporting that comes out … that’s not going to represent who the people will actually want to be the leader—something in that, will create so much division, that people will resort to real violence; that the United States loses its “united” aspect. So evil people, evil forces… just cruel human beings, use that moment to attain power. The kind of power that goes against the ideal of the United States. It could be Donald Trump; it could be another human being. It doesn’t matter; the point is that love doesn’t win out. .. I feel that you and I have a responsibility; how do we let love win in this moment of potential chaos
E: We’re going to have to fight for it. You’re going to have to throw some serious punches
L: Have to be Ali; the moment you start criticizing anything; you have to be a masterful communicator
E: That’s why you’re here! Look, Lex—in part, your decency is allowing you to do things you couldn’t otherwise do. I saw you had Michael Malice on; he is, I think of somebody who at his best, is extremely shrewd and insightful. He’s also got this trolling game, which I can’t stand. “Oh, grandpa doesn’t get the Internet.” … There are trolls of the past, who were incredibly good. I don’t see any of the modern trolls as being that kind of genius level, trolling the people who deserve it, in the way they deserve it. Right now, I see anything that stands up, gets cut down. .. The people who believe the world is chaos and has no point, always want to let you know, don’t try to use the Internet for meaning, decency, goodness, because we are going to find out that that’s all sanctimonious hypocrisy … There is a lot of sanctimonious hypocrisy in the world … The trolls somewhat remove that. It’s not a judicious, kind, compassionate, caring version. .. I have this feeling about Michael Malice; there’s someone that deeply cares beneath that. You and I don’t seem to be doing that; you’re almost never trolling. I’m very much against trolling. .. That’s not even true. Everything we say; I’m for, I’m against. This isn’t my native language. I speak nuance; I don’t speak this Internet shit. The more I have to communicate through Internet shit; I almost never take a tweet seriously
L: There’s an interesting effect where people say stuff and finish with “LOL” … it indicates to me … I wear this suit to fight the “LOl” at the end of sentences. It’s like, stand up for the words you’re saying. Don’t finish with “LOL” to remove responsibility
E: Also, good to choose the outfits of MIB and the Blues Brothers. .. But you need to be seated in a different chair. You’re in the wrong chair. It’s been so long; I want to talk about you and Joe Biden. Joe Biden was a 29-year-old guy with nothing particular going on, so far as I could tell. I know people as impressive at age 29 as Joe Biden—a huge number of them. None of them, my age, can get where he got. Any time somebody takes out; like if you found Eddie Van Halen in a guitar shop, you’d be angry. What is this guy doing repairing guitars? “Maybe he loves that.”
L: Yeah, what about your Russian piano tuner?
E: That was the point of that story. What happened in that life that converted somebody; I find this with Russian doctors that have become technicians. There’s a huge amount of talent in the world not sitting in its proper seat. We’ve got to take the seats! Maybe we don’t sit in them; maybe we take them and put some smart Gen Z person in the seat. Put them there and say, “No chanting!” If it rhymes, it’s wrong! In general, if something starts out, “1, 2, 3, 4,” I don’t want to hear it!
L: I feel the responsibility
Joe Rogan
… E: Joe and I have become friends, although sometimes we have miscommunications. On Yom Kippur, I texted him; I said I want to apologize for ways I’ve let you down as a friend … He responded, “What the fuck is your problem? Dude have you lost your mind?” I was like, you don’t know about Yom Kippur
L: What do you think about the Spotify thing … You could be that too
E: The instant Joe takes an interest in politics, and saving the world—that’s going bye-bye.
L: I disagree with you!
E: Okay; I bet you a bottle of Stolle; if you get Joe to call out the system for the bullshit that it is, and he doesn’t get destroyed, I’ll give you the vodka. No living heroes, my friend.
L: It’s just difficult; you just have to be good at it. ..
E: The more heroic you are, the more beautiful you are, the more you will be made to suffer. … If they cannot get you on reputation .. if Jesus himself came down! I probably never read for you the hit piece I did on Jesus … I did hit pieces on all the best people in the world; to prove I could do it to anybody at any time. Eddie Van Halen is now dead. But you know, “Hot for teacher?” Cancelled! Also, packaging female objectification for young men? Cleary, one of the worst people alive.
L: But the inspiration that is radiating from his music, that they will fight this cancellation
E: This is your thing; that in good versus evil, good has already been designated the winner. … Like I also believe that we’re going to win. Otherwise I can’t get out of bed. It’s pretty heavy at the moment.
Our political leaders
L: Do you think 2021 can make us feel good about the trajectory? Like there’s a smile on Eric Weinstein’s face? We’ll be doing a duet in guitar, not having this worried look on our faces?
E: No.
L: You’ve also promised that we’ll end this on a positive
E: Until we get some actually decent people in the right chairs, who are not constantly thinking about the next paycheck. The prerequisites for a solution, and why I don’t think it’s coming. Both of these political parties, the leadership is disgusting, and has to go. They lack the will to be Americans. They’re simply the people who have figured out how to inhabit the seats; that is their great achievement. You need people who can integrate and are not partisan at the level … People who believe in dividing the pie of the future as opposed to the present pie. … You need an ability to have subtle conversations; and you need the ability to exclude. Right now, everyone says “inclusion is good.” Like saying “water is good.” But it’s not! People drown. Inclusion is just inclusion; it’s not good or bad. We’ve taught people they can reason through the world as sub-cocker spaniels; they just bark things at each other. “I’m for inclusion! I’m for safety!” Are there complex sentences? … Where are we in this sea of nonsense. You have to be able to have smart, talented people who represent a diverse group of smart opinions. You have to get rid of all of the people who have the opinions that are antithetical … You need to give them insulation; we’re terrified, we don’t trust anybody; we want all walls to be plexiglass. We have too much transparency, and not enough at the same time. You need to ensure that people aren’t worried about feeding their family, for being real! And our billionaires? They’re pathetic; what’s the point of being that if you’re not going to do billionaire-type things. What is the point of creating obscene wealth if we don’t have anyone smart enough and caring enough to use it.
L: Let me slightly push back, on the idea the leaders themselves are broken. This goes to the Joe Rogan, Joe Biden and Trump, having a debate on that program. I feel like Joe Biden has a lot of interesting ideas, that he’s almost forgot how to communicate. He’s been fake for so long in the system
E: Hillary was fake for too long
L: If the platforms empowered you to search for those ideas within yourself, like long-form conversations do, then even the leaders we have now would take this country to a better place that would unite people. We can keep the current Congress; … this goes to … thee’s good in Donald Trump
E: There is good in Donald Trump. There’s good in Joe Biden
L: Several things are broken. One is Twitter. Another is journalism … just platforms
Younger people should have more power
E: I try to come up with unifying explanations. If you look at wildfires in California; … it looks like wack-a-mole .. you want to … I found a unifying explanation that really speaks to me; people are frustrated … Here’s the graph you need to look at. On the x-axis is time, by year. And on the y-axis, is something like average age of a human. The title of the graph is, Any Desirable Situation Involving Institutions. That could be CEO, tenured professor, grant-getting … All these things go up. In other words, for a long period of time, the average age of the person in a desirable situation has been increasing. Those graphs have to go down at some point. The specter of having five people all born in the 1940s as the final entrants in the presidential contest. That makes no sense. The last five people were all ancient, by presidential standards. We are talking about a contest between somebody who is the oldest of the baby boomers, fighting somebody who is in the Silent Generation. The Silent Generation guy gets a question from a Generation Z guy … It’s bizarre! It’s preposterous. That graph is the graph we can’t talk about; it is the graph of our destruction. Because, you can make a one-line argument: “Sounds like age-ism!” That muddles the conversation. Who wins as a result of the muddling?
L: It’s a battle! Let’s win it; let’s win the battle. … I was born in Russia; I can’t run … this is me officially announcing …
E: Lex, what do you really want to ask?
L: I want to put some responsibility on the Portal; that the Portal gives power to the people on that graph. You put it quite brilliantly that the people who move the world, their age has been going up. Not move the world, but put in a position where they get the chance to affect the world. These new platforms, like Twitter, give … to the younger people. You, you’re like Gandolf; you get to pick your Frodos.
E: I don’t know if people know where to fit me into this ecosystem
L: Well, figure it out! Maybe you need a mustache; I don’t know. There’s something about figuring out how to be a charismatic communicator. …
The Portal
E: … This election is chewing up the integrity of everyone who comments on it, Lex.
L: Maybe they’re not good enough. The hope is, you’re good enough.
E: Listen to me very careful. My spider sense tells me … this is a super-dangerous time; for smart people to be spending the dry powder, because the election doesn’t make sense. One outcome would be better than the other, probably … these two options are so completely inappropriate to the world of 2020. What we need is so diametrically opposed to more Boomers and more Silent Generation people trying to sort out a highly technical people being mediated through social media … we need more exclusion; we need more actual elites.
L: Yeah, we need excellence, competence
E: We need people who can be trusted behind closed doors, and we need to close the doors … Imagine you had a bunch of people who had all seen action in combat; had all volunteered; had all come from backgrounds where they didn’t need to … imagine you had 10 of these people with technical backgrounds. Men, women, black white, Muslim, Jew, doesn’t matter. I would trust those people, and I would close the door.
L: I disagree with you there; there’s a difference between those and Jocko. … The powerful thing about Jocko, he’s not only self-credential, but he’s been real. The magical thing … he’s been talking on a podcast for a long—there’s something real that happens
E: If you took Dan Crenshaw, Tulsi Gabbard, maybe Jocko, … take Bernie Sanders, a lone voice. All these people who really risked. Why is Katherine Hepburn the best Hollywood ever produced? Because she told Hollywood to go fuck itself. …
L: That’s skill
E: That’s what you were talking about. Be Katherine Hepburn. Audrey Hepburn is pretty amazing, but Katherine Hepburn is next-level … I don’t have the answer … The parties don’t have enough integrity. If you comment, either for or against; you make a comment that’s nuanced, you get referenced to something … Take this thing, “Fine people on both sides.” That is non-resolved after n years, whether the context should be reported or not. … Democrats and Republicans are primed to fight each other … I don’t want to be in that fray, because those people are going to kill each other, mindlessly, like robots. Until the election is concluded … do I think this is dire? Yes. … I can make an excellent argument that we need to elect Joe Biden now. I could make another argument … that .. can only be cured by defeating Joe Biden right now .. and all of the things the modern Democratic party represents. It’s the tiger and the tiger. Which tiger do I have a better chance against? The key problem, politically, is we have to divorce the concept of the center and moderation from kleptocracy. Whenever we say more moderate solutions, the reaction is, you want to hand this right back to the swamp, don’t you? So, we have these two crazy wings … We need an entirely different system. People say, what you’re going to sit this one out—because you’re a coward!
L: But are you thinking of what to do
E: Absolutely. Brett had this idea of Unity 2020 and I told him it was the wrong idea. … If I were to make the case that he was right, I would say the case to be made that he was correct; by doing it in 2020, we found out what we were up against. We learned that Twitter could turn this off at the drop of a hat. It’s good to know we cannot have meetings of presidential candidates in a primary that are not approved of by the party. Now Unity 2024 makes sense because Unity 2020 was tried. I don’t know that we get to 2024 .. in all circumstances
L: There’s a game theoretic thing that … Jack Dorsey, very likely, listens to your podcast. This is the power of these words
E: Something deep went wrong
L: But, we can change it, with the power of words
E: Something went wrong at Twitter
L: They’ve got so much division at their platform. It’s not wrong; it’s understaffed.
E: They’ve got an unsolvable problem
L: … disagree. I’d like to create a competitor to Twitter.
E: I can tell you things I’ve talked to Jack about that I know would make Twitter much better … The problem is us. It’s not the platforms. You’re thinking about a technological solution; I’m saying the problem is that we are ultimately the product.
L: I disagree with that …
E: I look forward to spending summers in your villa when you create this product …
Money
L: I will never have a villa; I’ll always give away everything I own
E; Don’t do that. Don’t pledge to be the person who disgorges themselves of security. Money is freedom; that’s what it is. It’s a big hunking pile of freedom. You could choose to use it as the … to imprison you. .. But generally speaking, money is freedom. … Your voice is important. AT least, retain the amount of security you need to follow Joe’s advice. What’s the point of F you money, if you don’t say F you … Others built a prison with the freedom they had …
L: The reason I want to give away the money, is I know my own psychology, and you create prisons … The optimal F You …
E: Do you have kids? You don’t have kids yet.
L; This is me, single Lex …
E: I’m talking to future Lex. Single present Lex; don’t be an ass. You need money, to do many of the beautiful things we’re counting on you to do. Don’t F it up.
Roger Penrose
L: … he just won the Nobel Prize for physics. He had a connection with you … clearly somebody profound in your world view (like Eddie Van Halen) … what does it mean that he won this highest of prizes
E: First of all, there were two other people who won; but I didn’t know them before they won. IT’s not Roger in particular, but the class from which he comes that is so important. I would put him in the class of Feinman-Einstein-Dirac-Yang …
L: Put Whitman there, or no?
E: Yeah. Witman’s a special case; he’s weirdly the reverse of the Penrose story; he’s the first physicist to win a Math Fields medal. Penrose is a mathematician who’s now won the Nobel Prize. Roger’s class means everything to me. That’s the highest achievement of the human mind. I’d probably throw Bach in with them … I think that he was so inventive; it was very frustrating to watch his career (same with Feinman). Feinman was so good, and had he been born slightly different, at a slightly different time, I believe his claim on physics would be far greater. I feel like Penrose … came up a very difficult path. Einstein effectively solved many of the important problems of general relativity in the beginning. Thus the children of Einstein are impoverished; there wasn’t as much to pick off the trees … I think Roger affected me personally, by a diagram that I saw in a paper of Herman Gluck at UPenn; it was the first picture I’d ever seen of the Hopp Vibration sketched. Weirdly, I brought that to the Rogan program to explain the wonder; I think I saw that at age 16; it just flipped my mind. Roger’s incredibly visual, geometric, sui generis. He does his own thing. He’s got lots of bets; none of which came through the way you would hope. I think they stretched the rules to give him the prize.
L: You said on Twitter; every once in a while, a human comes along that gives value to the prize, rather than the prize giving value to the human
E: We don’t care about the prize because of Alfred Nobel. We care about it because it came along at the right time to honor Einstein, Dirac, Feinman … You should have a love-hate relationship with it. On the one hand it focuses the world on what really matters, but it distorts what really matters. In this case, I think they violated their own rules, slightly. … … make sure the prize is funded fully going forward. Roger is getting on in years; the person should be alive. I think they bent the rules; they couldn’t have done it for a better person. .. It would have been great to get Madame Wu a posthumous prize along with __, ___, ___ … and George Zweig … there have been some terrible omissions. I take a dim view for people pushing for prizes from ethnic groups … or genders, if it’s not following the work. But there was something wrong with the Nobel committee … try going through the day without using ___ Theorem, and without Madame Wu’s contribution
L: Like you said, the Nobel Prize is plagued by omissions
E: And distortions and dilutions. Dirac and Schrodinger were given the prize in the same year; there’s no reason those two should have diluted each other … I don’t think we needed to dilute Weinberg, or Feinman, or Schwimer. It makes me sick; they’re such important giants. I think it has to do with the field not wanting to create luminaries or superstars, that could have defended the field from budget cuts … It’s important that we produce superstars … and important that the prize stays funded with the prestige … that comes from giving to types like Roger Penrose.
Jeffrey Epstein
L: Can I talk to you about evil? … Everybody at MIT is quiet about it. Which is Jeffrey Epstein. I didn’t get to experience what MIT was like at the time Epstein was part of it. I’d love to try to understand how evil was allowed to flourish in the place that I love. Whether you think; let me ask the question this way. Was the man evil, or was the system evil? Or is evil too strong a word? Because what I see is the presence of this particular human being … in the eyes of many destroyed the reputations of many really strong scientists. And also weakened the institution of MIT, by making everybody quiet. Almost making them unable to say anything interesting or difficult. … What is that? And what am I supposed to …
E: We don’t know. We don’t know. I obviously I want to scream about it too. I probably have said too much about JE. Something horrible happened; I don’t know what it is. Something horrible happened; and, one thing that—The first thing I need to do is get rid of this woke crap about power differentials. In general, you can talk about hypergamy and powerful differentials; those are Ruissell conjugates of the same concept; just like particular proportions and symmetries are mathematically provable to be attractive in females to maels … male attractiveness is largely determined by competence and ability to amass power and success. The relationship between consenting adults is frankly not something I want to sort out. The relationship between the sexuality of adults and minors, and particularly … The 17, 18 issue—that’s very different than 12, 13. We’re talking about really sick depravity, with respect to what it appears JE was involved in at some level. I believe this story is super-complicated in part, becasue I think one thing he was doing was providing money, encouragement, and support to scientists. Another thing he was doing was giving tax advice to very rich people. Another thing he was doing was hooking very wealthy people up with young adult females. Another thing he was doing, I think, was doing stuff with children that wwill curl your toes. There’s an entire spectrum of different stuff; nobody can pull apart or de-confalte anything, because the woke thing comes over it, and says, “I think ti’s disgusting that a 43-year-old billionaire would be partying with a 23-year-old.” I don’t want to adjudicate that … I don’t think MIT was deep into pedophilia. .. I don’t think the scientists were. the targets of the really sick depraved stuff. What you’re looking at was a government construct. Maybe ours; maybe a joint government project; maybe somebody else’s. I believe that in part, we don’tt really understand Robert Maxwell … Gelaine Maxwwell’s father, very active in scietniffic publishing. I don’t know where peer review came from; its relationship to Robert Maxwell … I would love to run down the missing fortune of Robert Maxwell, and the mysterious fortune of JE, because I don’t think JE ever ran a hedge fund or was a money advisor.
L: Two things. One, the shallow conversations of woke identity politics, seems to be removing everyone’s abiliity, or wililngness, to talk about what the hell is this person? And how is he allowed? And how do we prevent it in the future? And second, from the individual perspective, it’s the same question as from 1930s Nazi Germany—if I was there at that time, what is the heroic action to take? Same with JE; what is the heroic action?
E: You wouldn’t know; you don’t know what you’re up against
L: That question might be the heroic action
E: You have to map the silence with JE> What you’re describing is a map of the silence of JE. What about in the state of Washington, the Bay Area, NYC. The amount of silence around him should be telling you everything. The number of dogs that don’t bark is .. al ot
L: What it’s telling me is not some kind of consipracy, but more a disappointing weakness …
E: Wait, not some kind of conpspiracy! You’ve got to be kidding! You’re so afraid of saying the word conspiracy
L: I just think it’s people I thought were my heroes just being weak
E: No. Be of good cheer, sir. I think there is a conspiracy.
L: That would be a very impressive one. I tend to think large-scale can only be an emergent phenomenon. I find this fascinating; because you’re very logical; logic and love drive your soul. I believe if you would review the video of the mysterious hit on the hotel guest … the assassination, in 2010, of Mahmoud al-Mabu, something like that, in Dubai, where I believe 26 separate individuals, on multiple teams, are shown converging, from all over the world, on false passports, pretending to be tennis players, or business people, or vacationers … all of these teams have different functions, and they murder this guy in this hotel room. The Dubai chief of police was so angered, that he put together this amazing video, that says, we can completely detail what you did; we caught you on closed circuit TV; we don’t know exactly who you are … No! I don’t believe, after COINTELPRO, and Operation Paperclip, and Operation Mocking bird, and maybe Rex-84, to not believe in conspiracies is an idiocy.
L: So you have a sense that evil can be as competent, or more competent
E: When evil wants to operate at scale, it needs to make sure people don’t try to figure it out. When it operates at scale, from first principles, evil must not want it investigated. The most efficient way to keep yourself from being investigated .. is to invest in a world in which no one can afford to say the word conspiracy. There’s a special radioactivity around the word. We have provable conspiracies … you have been invited to conspiracy; they are everywhere. Some are mundane; some are price-fixing cartels, or trade groups. The first thing you have to realize, is all of us are under a memetic complex, where you can be taken off the chess board for saying, conspiracy theory. … That is partially distorting our conversation; if you want to ask me about JE, you have to agree with me that that’s a logical description of what you’d have to have if you wanted to commit conspiracies.
L: It’s a fascinatingly difficult idea. The world with conspiracy theories, and the world without conspiracies .. look the same.
E: Well, there’s responsible conspiracy theorizing. Just as you would with any other topic; think how different the rules are for conspiracy theorizing versus X-theorizing. If I say to you, average weight is not the same between widely separated populations. … In fact, no continuous variable that shows variation should be expected to be the same between widely separated populations. Yeah. Like IQ. Wait! Hold on! So the first thing to notice is that conspiracy has that baked into everyone’s mind.
L: That’s really important to say. Yeah. And that would be the first step, if you wanted to pull off a conspiracy in a competent way, you would have to convince the world ….
E: I just watched the movie 1971 … a citizen’s committee that investigated the FBI; they broke into FBI offices to unearth a conspiracy inside the federal government. It’s a double conspiracy story. I think the problem with modern Americans is that they are so timid, they don’t even learn about the history of conspiracies that are absolutely proven. So, JE represented somebody’s construction.
L: Kind of scary.
E: Yeah, what part isn’t scary? … I was scared about having a conversation about JE; I was trying to let it be known that I don’t know anything more than I’ve already said. Now your friends at MIT; their problem is, JE showed up as the only person capable of continuing US scientific tradition. IT’s a little like the Russian; it’s combatitive. We’re a free society, and we act like a free society; we’re rich and we research like we’re rich. Then came the 1970s, and the Golden Fleece awards, and the idea these are welfare queens in lab coats. Everything went to hell. The national culture of U.S. science was lost; and then JE shows up, and a tiny number of funders: maybe Fred Kavley, maybe Yuri Milner, maybe .. who else? Peter Thiel to an extent. Howard Hughes would be the largest of these things; different grant structures … gave people a modicum of risk-taking ability. When JE showed up; everybody wanted to take risk in science. Suddenly a charismatic billionaire said, here’s a (bunch of money)! That money was supposed to be provided under the Endless Frontier compact between the federal government and the universities. The federal government welched. … So the universities became psychopathic. There wasn’t enough money to be moral; so it was time to eye each other as a source of protein, as I like to say. JE said, hey, come to my world. We can do it like we used to do. In part, none of your colleagues at MIT, had that kind of religious commitment to science, that they’re willing to go down with ship science. The Galileo Galilei thing became important, because sometimes you have to say, this isn’t about me and you. There isn’t enough money in the world to buy the legacy I want to leave on this planet. That’s one of the good things about it—it’s worth dying for … I’m not eager to martyr myself, but I’ve certainly risked my health, my fortune … I’ve destroyed myself economically over science. … My need to oppose these sons of bitches in chaired professorships who are destroying our system …
We need to get off this planet
L: Let me bring Grandmaster Ugwe into this … Are these Kung Fu Panda references? Alright, you’re not supposed to call out my journalistic integrity. Master Ugwe says a couple things I’d like to bring up with you. One, he recommends that you should find a battle worth fighting. We’ve talked about several battles just now. What is the battle worth fighting for, for EW, in the next few months, in the next year.
E: There’s only one. It’s time to go. It’s time to leave. It’s the Moses thing. Look at your world! You just got introduced to the problem of a virus. Wait till it’s fusion devices; and you have one interconnected planet with no … experiments happening anywhere else … Everybody who has a possible plan to avoid what is coming if we don’t have one, should work on the plan that he/she thinks best. Elon wants to do rockets. People misinterpret me. … Meta Eric says, I don’t think that’s a small plan. But REgular Eric says, all people who have hope should do that thing. At least it’s Mars.
L: It’s the kind of fight worth fighting.
E: Same as Brett’s Unity 2020. Everybody should do something. … I’m hopeful about it. .. If I thought Daniel Schmactenberger’s Wisdom Project was a better hope, I’d do that. Look, we got from powered flight with the Wright brothers to sending back photos from the moon in less than a century. If we can change the laws of physics … it may be that it buys us nothing. But at least we will know why we died on this planet.
L: As a small aside, I feel like you will guide me, like Master Ugwe did; and I’m the Kung Fu Panda.
E: They only had one conversations.
An update on Geometric Unity\
L: Well, we’re Jews, and they’re not, so we talk too much …. that you would guide me through some more intuition on the source code of our universe. Where are you at with Geometric Unity?
E: I’m trying to figure out where to release it and how! … The video’s coming up on half a million views on YouTube alone. To say nothing of the audio. .. It produced a strange reaction. One thing I don’t think I possibly understood, is most physicists don’t talk in this geometric language. I thought more of the physics world had converted over to manifolds, bundles, … I saw a lot of the comments; people with PhDs in theoretical physics saying, I’m not familiar with these concepts.
L: What’s the solution to that?
E: My problem is … my calculation is that as long as the boomers are still in charge, the same people have these incentives, where they’ve invested in programs that didn’t work, so they’re extremely hostile … I think you had Wolfram on your program; I don’t remember if he said this to you or Brian Keating—everyone got discouraged. IT was too hard. Something about the re-normalization revolution (energized) the physics community … somehow that discouraged people from guessing, believing … Renormalization was one of the most important revolutions that happened in science. The story that the physics community told about it was casastrophic.
L: The podcast medium is revolutionary for discourse. Are you thinking of revolutionary ideas for re-energizing the physics community?
E: I have a fantasy; all of these things are the same problem. We read about these things in Feinman’s books. What’s the greatest thing that ever happened in math? Tartaglia’s solution of the cubic. Why? It was the first time a modern person had done something profound that the ancients had failed to do. So it opens up psychology that maybe things are possible again. New farmers, who can find fruit that they can pick … The reason he created a revolution … you could follow Eddie Van Halen. … You could follow … Once you understood there is a tapping principal, it was … percussive guitar. Once we start innovating in the present, everything will come. Because everything around us is screwed up.
Gratitude
L: Yesterdays’ history, tomorrow’s a mystery, but today is a gift. That’s why we call it the present. … (Another Kung Fu panda quote). Let me ask, what are you grateful for today; what is your present. We’ve talked about a lot of dark things, but what brings joy to your heart …
E: Nyla and Zed, my wife Pie, the fact that weve got our health, the little things. Saying grace after meals. You’re coming over for Friday night Shabbat dinner; we’ll bench together and say grace. This bottle of water in front of me; I made a point of thinking how wonderful it is a quenching bottle is placed in front of me because somebody cares. I still have strength for the fight. I’m grateful for that; I can’t believe I’m not more beaten down. I have the most interesting set of friends; I really do. I’m not that rich, by monetary standards. If there were friend billionaires, Forbes would be all over my ass. I’m really grateful … I think this is the end of something profound, and it’s the beginning of whatever is next. Whatever’s next could be terminal; it could be amazing; it could be a return to the horrors of the 20th century … and takes millions of lives in the process. I’m grateful to have half my life in the rear-view mirror. Maybe it took place in a bubble, and was unsustainable. But it was nice to move around the world without a maskk.
L: To fall in love; to have a famil.
E: Absolutely—to find the last Indian Jewish girl left. … I forgot to say, falling in love with an intellectual collaborator is a great thing that not everybody has a chance to do. She and I had an adversarial relationship around geometry and economics. … Like a buddy picture, where the … we came up with geometric marginalism, this other theory, which allowed me to inhabit space with somebody who I already knew intimately and had fallen in love with. It’s sort of the intellectual version of the tango. … Version of my life that doesn’t fall into people’s experience. … She doesn’t want to revisit the material.
L: But to flip the tables on you … I do hope your voice continues to reverberate, at least in the 2021s, and beyond.
E: We’re pausing at the moment; we’ve recorded some for future episodes. … Earnestness trades at a discount at the moment because it’s easy to make fun of it. What I like about you is that we’re fairly earnest. I want our listeners to know that I’m not stepping away from the podcast; this is hugely financially costly to me. I want to make sure you guys are getting the best that I can do; and destroying myself right in front of the election; I think Lex is incorrect; I think the forces that are trying to make sure there aren’t any planes in the sky that aren’t red or blue … given how angry I am at the system … if nobody’s going to talk about JE, I want to make sure that I’m there. Do I think it’s potentially an existential election … yes. But I don’t know; Where we are right now seems so dumb, and castatrophic in how it is chewing up smart people. It’s not about cowardice; it’s hard for me to restrain myself … This is about trying to make sure I’m around. I was afraid what happened to Brett’s articles of unity was going to happen to Brett, was going to happen to a few of us. … The Intellectual Dark Web …. The thing is to be hard to kill. Because when the hit pieces come, they come for where they can get you. It’s important, right in front of an election; the desire of an old system to defend itself … through reputational destruction. … They have plenty of weaponry with which to fight us. My goal is to stay here as long as possible … Help other podcasters start, and my hope is that works; long heroism, short martyrdom is a good motto for anyone. I try to remember the short martyrdom part of that. …
L: Way to end the conversation on a disagreement, which is how you hook them for the next conversation, to be continued.
#135: Charles Isbell, G. Tech dean
Charles Isbell, dean of the College of Computing at Georgia Tech; a researcher and educator in the field of AI; and someone who deeply thinks about what exactly is the field of computing, and how do we teach it. He also has a fascinatingly varied set of interests, including books, movies, sports, and history—that make him especially fun to talk with.
I’m trying to make it so that conversations #134, 135, and 136 all are released before the Election. None of these are explicitly political; but I hope these conversations can “bring context to our difficult time.” With Eric, we talk about the nature of evil. With Charles, we talk a little about race in America. And with Dan Carlin—we talk about Alexander the Great, Ghengis Khan, Hitler, Stalin … and many other things in between.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
L: You have mentioned that you love movies and TV shows; let’s ask an easy question, but you must be conclusive. What’s your top three movies of all time. You have to be definitive and conclusive.
C: So it’s hard to be definitive and conclusive. Movies is too broad of a category. I’ll pick one or two subgenres. My favorite comedy of all time is His Girl Friday; it’s based on a play called The Front Page; the movie is a fantastic film. It’s a screwball comedy. You’ve seen Moonlighting, the TV show. A man and a woman, clearly in love, always talking over one another. This was the movie that started all that. It must have come out between 1934 and 1939. It’s black-and-white; it’s a fantastic film.
L: So it’s mostly conversation.
C: Not entirely, but mostly. There’s a story; someone’s on death row. They’re newspapermen, including her—they’re all newspapermen. They were divorced; the editor, and the publisher. He’s trying to get back together. But none of that matters; the plot doesn’t matter. At the end of the day, narrative and conversation are things that drive me. Now I’ll cheat and give two movies as one. Crouching Tiger, Hidden Dragon and John Wick. Both recent. I love all the John Wick movies. I put them together because I actually think they’re the same movie. Both of them create a world that you’re coming in the middle of, and they don’t explain it to you. BUt the story is done so well that you pick it up. Apparently every person in NYC is an assassin. It’s a complicated world, and everyone knows each other. You get the feeling this is Chapter 9 of a 10-part story.
L: Martial arts—you have a long list of hobbies, but I don’t see martial arts as one of them.
C: I don’t do martial arts but I certainly watch them …
L: You like Rush Hour, too?
C: Sure. My favorite Jackie Chan movie would be Drunken Master 2.
L: He creates his own martial art … was he actually drinking? Or play-drinking?
C: You mean, as an actor? No, he was definitely drinking. In the end he drinks industrial-grade alcohol, and has one of the most fantastic fights ever. But I’ll tell you the last movie; it’s called Nothing But A Man, it’s the 1960s, it stars Ivan Dixon, who you’ll know from Hogan’s Heroes. A small little drama; a beautiful story. One of my favorite movies just for the ending is The Godfather; I think the last scene is fantastic; the whole movie is summarized in the last eight, nine seconds. It ends with the wife coming to Michael. He says just this one time I’ll tell you my business; she asks, did you do this terrible thing. And he says no. And then she’s leaving … and all these people are coming and kissing his ring.
People are easily predictable
L: You did an experiment where you tracked all this information about yourself, and a few others; sort of wiring up your home. In that video, you mentioned this idea, that two days worth of data is enough to capture the majority of the behavior of the human being. Can you describe what you did to collect all the data, and also what the intuition behind that is.
C: I was thinking specifically—I’ve been a part of a suite of these experiments. The thing I was talking about had to do with …. recording all the infrared going on in my house. This was a long time ago … I was trying to figure if we could get enough data on people to figure what they were gonna do with their TV or their lights. … It was kind of surprising; it shouldn’t have been. It’s very easy to do; a bunch of computer systems … So, it turns out that, and I did this with myself and then with students. AT the end of the day, people do the same things over and over and over again. IT has to be the right two days, like a weekend. Not only can you predict what they’re going to do next, but you can do it with something really, really simple. … It turns out you can get 93% accuracy … What’s more interesting, is you can use that information. If you try to represent people or objects by the things they do, the things you can measure about them measuring action in the world … you try to represent them … People cluster remarkably well, in fact irritatingly so. By clustering people this way … I got the … collection of things you might press. …
Breaking out of our bubbles
L: … we have these silos in social media; we have this clustering. Along this narrative, we want to break each other out of these silos so we can be empathetic to other people. … Democrats and Republicans are just too silly bins we seem to be very excited about. From an artificial intelligence perspective … throwing agents into that mix; interacting with us humans, getting us out of those silos; do you think that’s possible? A hopeful possibility for AI systems? To get us outside of those habits in the idea space? So we can be empathetic to other people’s lived experiences?
C: Yes. I don’t think it’s that hard—in this sense. Imagine that you can … let’s assume that you can do a kind of partial ordering over ideas, or clusterings of behavior. Two clusters, some edge between them. If you can imagine that, then the way you get from here to here is you find the edge, and move slowly together. I think that machines are actually very good at that sort of thing. If you have the network and you know the relationship, you have some semantic meaning for them—then I think you can kind of move people along. But it’s harder than that for this reason: I’m going to tell you a story that someone else told me: You take two sets of people from the same backgrounds, and you want them to solve a problem. You separate them, and then you have them in this big room, far apart from one another. When they come back to talk about what they learned, you want to merge them together. IT can be hard, because they don’t speak the same language anymore. The example I was given—they’re sitting over there, talking about these rooms you can see; you’re seeing them from different vantage points. They start referring to the room as “the one with the clock.” They end up referring to it … they’re referring to the same room … the clock on the wall is the thing that stuck with me. The problem isn’t that the ideologies disagree. The problem is you’re using the same words meaning radically different things. Get them to agree on some basic definitions. Right now, they’re talking past each other. Getting them to meet and interact might not be that difficult. …
L: It’s an interesting question to me … it can be a worldview. It all boils down to empathy; being willing to put your feet in the shoes of another person. To learn the language, how they see the world. I experience this now with trolls. I talk about love; I’m lucky to have this community of loving people. But whenever I encounter trolls, they roll their eyes at the idea of love; they call it cringe. They show love by derision … I wonder if AI systems can help somehow. Bridge the gap of, what is this person’s life like? Encourage me to ask that question; to put myself in their shoes; to experience the agitations, fears, hopes they have. To think about what was their upbringing life? A single-parent home? A shitty education? To put myself in that mindspace … It feels really important for us to bring those clusters together. It would seem AI needs to understand both parties first.
C: Do you have to understand it or do you simply have to note that there is something similar; a point to touch? The word “understand” there is doing a lot of work. You use the word empathy; I like that. Let’s separate that from sympathy. Sympathy is feeling sort of for someone; empathy is understanding where they’re coming from and how they feel, right? For most of people those things go hand-in-hand; some are good at empathy and bad at sympathy (or vice versa)? Some are incapable of feeling sympathy unless they feel empathy first. …
Interactive AI
Lifelong machine learning
Faculty hiring
University rankings
Science communicators
L: … there’s a litlte bit of jealousy. Whoever is popular, being a good communicator, exciting the world with their science. Of course, that’s not peer-reviewed, clean … It sounds like bullshit; like a TED talk. People roll their eyes; they hate that the TED talk gets millions of views. So they pull each other back. It’s only when you get senior enough that you stop giving a damn. That was always the surprising thing to me. Even when you get tenure … … ..
R: I’d be interested in how Rod Brooks felt about how people treated him when he was doing Fast, Cheap, and Out of Control … naked mall rats
L: Rob Brooks used to be head of the AI laboratory at MIT; then he launched “Rethink Robotics.” He’s a little bit of a rock star personality in the AI world.
R: Also, he was one of my two advisors for my PhD. I love Rod. Also, love Paul. … … He made a lot of his bones by doing counterintuitive science. He did amazing things, and continues to do those sorts of things. He might tell you …
Hip hop
L: This topic, I’m on thin ice. I grew up in the Soviet Union. Your research group is called …. but also, there’s a bunch of mystery around this: it’s also called P-Funk. P stands for Probabilistic; but what does “funk” stand for
R: A lot of my life is about making acronyms. .. PFUNK has three or four different meanings. At the end of the day I decided the P stands for probabilistic; the FUNK could be lots of different things; I decided it should be up to the individual. When my students graduate, I hand them a hat, star glasses, a medallion from the PFUNK era, and I hand them a pair of fuzzy dice.
L: There’s a sense to it that’s not an acronym—literally funk. You have a past … in hip-hop, and funk. So, can you educate a Soviet-born Russian about this thing called hip-hop. If you were to educate me, especially about the last couple of decades, what records or artists would you introduce me to?
C: All great music was made when I was 14 … Hip-hop and rap aren’t the same thing. Hip-hop is a culture, of which rap is a part. So tagging is a part of hip hop. … There’s popping and locking, and the dancing. It’s a way of life, which I think is true.
L: Rap is the music part.
C: It’s “a” music part. What rap albums best touch it? … I’d try to figure out what you like and work you there …
L: Lynrd Skynrd. Led Zeppelin.
C: There’s a fascinating exercise one can do, watching old episodes of “I love the 70’s,” “I love the 80’s,” “I love the 90’s” … I would start you with Public Enemy, particularly It takes a Nation of Millions to Hold Us Back. Fantastic lyrics; for me it’s all about the lyrics. Rick Rubin was the producer of that. And it was focused on politics in the 1980s, which was what made hip hop so great then. Then I would move you up through things happening more recently. I’d include Mos Def.
L: He hosted a poetry thing on HBO
C: I’d work you back to EPMD, and eventually to The Last Poets, which was 1970—to give you a sense of history, how it’s been building up over a very very long time. Then we’d cycle out, move back to the present, and then go back to the past. I think people are confused; it’s about narrative and being a part of something. Jazz, which I also like; one of the things that’s cool about it. You’ve been so immersed in it … I would help you to see that there’s a history of it. How it connects with other genres of music you might like. Including, some of the good work that’s been done with fusions of hip hop and bluegrass.
L: Oh no!
C: Yes! Some of it’s even good. …
L: There’s a tradition of integrating classic rock songs into hip hop beats. It’s kind of interesting. Not just classic rock; what is it, Kanye, Golddigger …
C: That’s been true since the beginning. That’s why the DJ used to get top billing; the DJ brought all the records together; made it so people could dance. The DJ showed music is itself an instrument. Very meta. That’s going way back. In the period of time when I grew up, which was mostly the 80’s … … … that’s probably because it’s being sampled by someone referring to something they remember from when they were young. This stuff’s been going on for a long time. Jammaster J played piano; he would record himself playing piano and then sample that; …
L; That’s how his mind …
C: It’s putting pieces together to build music …. but still, it’s the right attitude.
L: You mentioned lyrics; it does make me sad, this is me talking trash about modern hip hop. The lyrics went away from talking about, politics, life and so on, the protest songs, like a Bob Marley, or you said Public Enemy or Rage Against the Machine; that’s the place we go to those lyrics. Classic rock is all about, my woman left me, or I’m happy she’s still with me. It’s less interesting I would say. It seems like rap is the place you would find that. It’s sad that, from what I see, mumble-rap or whatever, they’re moving away from that.
C: I’ve always been a fan of lyrics. If you look at my reviews; I wrote my last review, the month I graduated and got my PhD. I would always start with it’s all about the lyrics. Someone has already written in the comments that neither of us knows what we’re talking about, and it’s all in the underground hip hop. And that’s always true; I discover some underground hip hop song, and I’m happy and made whole again. I don’t listen enough, because I’m listening to podcasts …
Funk
L: James Brown is funk, or no?
C: Yes
L: Little Richard? Ray Charles? Hit the road, Jack?
C: There’s definitely a funkiness in it; there’s a line that carries it all together. … It’s probably different to answer it in 2020 than it was in 1960. … I don’t think we used the word “Funk” in 1960.
L: Do you reject disco?
C: No, I appreciate all the mistakes that we made.
L: John Travolta; he rejects it probably.
Computing
C: How do you think about the world of computing; where it sits in the sets of different disciplines. How should people think about computing from an educational perspective; what is the perfect curriculum that defines for a young mind what computing is.
L: That’s an important question. Curriculum to me is the fundamental data structure …
C: I’ve spent a lot of time thinking about this. What does it mean to be a discipline? The truth is, we educate people; people talk about skill and skill sets; but disciplines are about mind sets; fundamental ways of thinking. It’s the mindset of the fundamental way to think about the world. Different disciplines give you different mindsets; different ways of thinking through. With that in mind, to ask about a discipline, you have to ask, does it have a mindset? A mathematician builds abstractions; tries to find steady-state truth about the abstractions. There’s the scientist, the engineer … What is computing? How is computing different? I’ve come to a view about what computing actually is … I think that, what distinguishes the computationalist from others, is he or she understands models and machines are equipment; they’re the same thing. It’s not just a model, it’s a machine that’s an executable thing, that can be described as a language—that means it’s dynamic. It’s like mathematics. The mathematician isn’t worried about the dynamic part. If I write something for mathematicians, they invariably demand that I make it static. That dynamic thing matters; that is what computing brought us. The models represent truths in the world … but of course it’s engineering, because you’re dealing with constraints in the world … but it’s also math … but the fact that .. regular expressions … that a language and a machine are essentially equivalent; that’s not a small thing. That idea is fundamental; we would do better if we made that explicit. … My life has changed in the 10-15 years since I tried to put that to paper … this sort of triangle of equality, it matters because there’s a person inside the triangle. What’s chnaged about computing, is we now have so much data, we’re able to do really really interesting, powerful things. It only matters with respect to human beings and their relationship to it. What makes it potentially world-species changing .. If the curriculum can convey that, then that’s a big win.
L: Do you pull human things, psychology, into the framework of computing? Philosophy? Studies of human nature?
C: BTW, it works both ways. … The cell is a bunch of if-then statements … Even the way we think about computation. The important thing to me … my engineering colleagues worry about computer science eating up engineering colleges. You shouldn’t worry about that at all. It’s central, but it’s not the most important thing in the world. … You’re not going to be a historian in 2030 without knowing something about data science and computing. You’re going to look at data … Same is true for psychology, a lot of these things. So the philosopher, the psychiatrist have a lot to say about computing, about the way people interact with computers. Certainly about intelligence.
L: Will computing eat other disciplines
C: some fields destroy themselves. One way is for the field to be everywhere .. you lose the thing that makes it worth studying. The thing about computing—although not unique to computing—is we are both a thriving major, in fact the thriving major, and we’re a service union, because people need to know the things we need to know. Our job, just as a mathematician’s job is to help someone think like a mathematician—our job is to help them think like a computationalist. We have to take those things seriously. We haven’t done so.
L: I just talked to Dan Carlin; he talks about this idea that it’s possible history as a field; currently most people study history a little bit. We kind of are aware of it; we have conversation … Most people have a curiosity and are able to learn it. His thought is, it’s possible given the way that social media works, that history becomes a niche field, where literally most people just ignore—everything’s happening so fast, and history starts losing its meaning. It’s like the theoretical computer science part of CS; a niche thing that only the rare person … It’s a kind of profound thing to think about, how we can lose these fields; in the case of history, it’s best for that to be a pervasive thing … I would think computing is similar to history in, it seems it should be a part of everybody’s life, to some degree, as we move into later parts of the 20th century. It’s not obvious that’s the way it will go. It might be in the hands of a few, still. It’s unclear that computer will win out. (versus Machine Learning?) You’re at the leadership level of this; you’re defining the future. … There’s this conversation of, everybody should learn to code; the changing nature of jobs. Do you have a sense of what your role in education of computing is, here? What’s the hopeful path forward.
C: First off, it would be an absolute shame if no one studied history. OTOH, as T approaches infinity, the amount of history is also growing, so you have to forget more and more of it. If you think of your brains being outside of your head, seems you could learn the history you need to know when you need to know it. …
L: It’s per our objective camera discussion
C: You have your own history. It’s even lost to you .
L: All the ex-girlfriends. Gone!
C: The big lessons of history shouldn’t be gone. The point is (with computing as well), you have to get across those lessons. You don’t want to lose the data … With computing, everyone doesn’t need to learn how to code; they need to learn to think in the way of being precise, in the sense of being repeatable. Being able to describe a problem in a way which is executable. Human beings are not very good at that. We keep talking back and forth to try and figure out what another person means. Having to think that precisely about things is quite important. That’s somewhat different from coding. Coding, a programming language, with a thing you fiddle with … One of the holes in ML is we forget we are doing software engineering. We’re using languages that express what we’re doing. We get so caught up in the deep network; we’re working with a set of parameters that we made up. So in CS education, we want students to be able to think like that. It’s a way of surfacing your assumptions; we call them parameters; they become if-then statements. … It follows from that that you have to be explicit about what you’re trying to do. You better get it right, or understand it … I think it’s key that we figure out how to educate everyone to think that way. Because at the end, it would not only make them better at whatever it is that they’re doing. It will also make them better citizens. It will help them understand what others are doing to them. You’re not going to solve the problem of social media, insofar as you think of social media as a problem, by making slightly better code …
Race
L: Let me try to proceed awkwardly into the subject of race. We’re living in a pretty tense time … in America. You grew up in Atlanta; born in Tennessee. You identify as an Atlanta native. You’ve mentioned that you grew up in a predominantly black neighborhood. …
C: I prefer Black—with a capital B.
L: (laughs) the rest of the letters
C: It doesn’t matter
L: And you didn’t almost see race? You can correct me on that … When you showed up to Georgia Tech, for your undergrad, you were one of the only black people there, and that was an experience …
C: That story continues through MIT—it was more stark there …
L: Georgia Tech was undergrad
C: Yes. I went straight from undergrad to grad school. There was nothing in between
L: You didn’t go backpacking in Europe
C: Anyway
L: Do you miss MIT?
C: I loved MIT. I don’t miss Boston—at all. Anyway, I was born in Chatenooga. Earliest memories are arriving in Atlanta at 3 and a half. … I loved it. Like much of the country, it was deeply, highly segregated, but not in a way that was obvious to you. But you could divide up Atlanta by highway, and get race and class that way. I grew up on the poor side of that predominantly black area. I was very much aware of race; my family made certain that I was. It would come up; in first grade, I had a girlfriend, a white girl named Heather. We had a long discussion about how it was okay for us to be boyfriend and girlfriend despite the fact that she was white and I was black. … That just meant we spent slightly more time together during recess. At the time it felt very scandalous—we may have done an Eskimo kiss
L: Did you write poetry?
C: No—this was first grade! So I was aware of it; didn’t think too much about it. … But I wasn’t aware that I was a minority. Because I wasn’t; as far as I was concerned. It didn’t feel that way at all. This being Atlanta, when I looked at TV, which back then one did, and when I saw the news, Monica Koffman was on TV telling me the news, and the mayor was black, and they were all black. When I went to Georgia Tech, the first day, walking across campus; I realized of the hundreds and hundreds of students I was seeing, I was the only black one. It was enlightening, and very off-putting. It continued that way for my career at Georgia Tech. I began to meet students of Asian descent, and those we would call Hispanic. That’s what college is supposed to do, open you up to people … When I came to Boston, I applied to one place as an undergrad—Georgia Tech. I didn’t know any better. No one told me. Grad school, I applied to GT, MIT, and CMU. When I got into MIT, I got into CMU, but I had a friend who went there; and he explained to me about Pittsburgh—something about the sun coming out two days out of the year. At MIT, I asked 20 people I knew, either when I visited or I had already known, whether they liked Boston. Ten loved it, ten hated it. The 10 who loved it were all white; the 10 who didn’t were all black. I could see it immediately …
L: The white people tell you about the nice coffee shops …
C: .. Harvard Square’s beautiful, something about the outdoors. The outdoors is for the bugs, it’s not for humans … the black folk told me different stories—which part of town you didn’t want to be caught in. I decided that MIT was a great place to be as a university, and that whatever it is I wanted to do, someone there would know how to do it … I thought I would be fine. I would only be there 4-5 years, I told myself. I did see a lot of—I ran across a lot of things that were driven by what I look like. I got asked a lot of questions; I ran into a lot of cops. At the time, these are the things I remember—this is 1990. There was not a single black radio station. This was 1990; I don’t know if there are any radio stations anymore. … But the idea of being in a major metropolitan area without a single black radio station, was absurd! I grew up in Atlanta; Boston had no economically viable, or cohesive black middle class. Insofar as it existed, it was uniformly distributed throughout large parts of the cities. Where you had concentrations of black Bostonians, they tended to be poor. I went to an eighth grade school that included a riot … but when I went to ninth grade, I went to Academy, a magnet. It was the first high school in the state of Georgia to sweep the state math and science fairs; it had 385 students, all but four of whom were black. I went to school with the daughter of the former mayor, Michael Jackson’s cousin
L: Dropping names!
C: Nine times removed, I don’t know. We had a parking problem, because the kids had cars. So, it was a very different experience … I had been to places where, whether you were rich or poor, you could be black or white. … It felt like a bunch of interesting contradictions. It felt like it was the interracial dating capital of the country. It also felt like the most racist place I’d ever spent any time. You couldn’t go up the Orange line; there were places you couldn’t go, and you knew it, everybody knew it. There were places you couldn’t live. That was the Greater Boston area in 1992.
L: Subtle or explicit racism?
C: both.
L: Was there levels in which you were empowered to be first, or one of the first, black people at some of these institutions? Was there part where it felt limiting?
C: I always felt empowered. Some of that was my own delusion. .. Not only did I not feel as if no one was trying to stop me; I had the distinct impression that people in power wanted me to succeed. I felt supported; at least that people were happy to see me succeed at least as much as anyone else. 1990 you were dealing with a different set of problems. You’re very early in the Jackie Robinson period. JR .. the first one has to be perfect. It was in everyone’s best interest; but I think it came from a sincere place. People went out of their way to make sure the environment would be good. … OTOH, we were 20 years away from the first Black PhD to graduate from MIT; Shirley Jackson, 1971, something like that. We were still eight years away from the first black PhD in CS. I didn’t feel as if the institutions of the university were against any of that; furthermore, I felt there was enough of a critical mass across the institute … who wanted to make certain that the right thing happened. Very different from institutions in the rest of the city; they felt no need to be supportive.
L: Let me ask a touchy question. You said that you didn’t feel … you felt empowered. Is there some lesson, advice, in the sense that, no matter what you should feel empowered? Is there a sense, from the individual perspective, where you should always, kind of, ignore the … uhh, ignore your own eyes?! Ignore the little forces you are able to observe around you, that are trying to mess with you; whether jealousy, hatred in its pure form, hatred in its deluded form? And kind of see yourself as empowered?
C: Sure, There’s a tradeoff right? You can’t get a PhD unless you think that you can invent something no one else has … But you can’t be so deluded … It helps to have a support group around you. I’ve been able to find that wherever I’ve gone … I had my mother and my family and those people back home, that I could always lean back on. Even if it was a long-distance call that cost money. You had that and it’s fine. But you can’t be so deluded that you miss the obvious; it will hurt you in the long run.
Cop story
L: You tell a story of being pulled over. Can you give me a sense of what the world looks like when the law doesn’t always look at you with a blank slate? With objective eyes?
C: It looks exactly the way it looks now; this is the world we have to live in; it’s people clustering, doing the things they do, making decisions based on one or two bits of information they find relevant … it cycles and cycles. It’s just about being on edge. I do not, despite having made it over 50 now
L: Congratulations, BTW!
C: I don’t imagine I will ever see a police officer and not get very very tense. Now, everyone gets a little tense! You’re probably being pulled over speeding, or you’re going to get a ticket. Most human being’s experience of the law is fundamentally negative. That’s just an annoying reality. … The time I got pulled over, halfway between Boston and Wellesley. I remember thinking, when he pulled his gun on me, that if he shot me, right now—he’d get away with it! It would be years later, when I realized, much worse than that—he’d get away with it, and if it became a thing other people knew about (odds were, it wouldn’t), not only would he get away with it, but I would be painted a villain. I was big and scary, and if only I hadn’t moved too fast. That hurts your family, your legacy. That’s terrible. And, it would work. Had he done it. He didn’t. He didn’t want to shoot me. … I wouldn’t be surprised if he never pulled his gun again …
L: You’re basically speeding or something
C: He said I ran a light; I don’t think I ran a light.
L: But he pulled a gun
C: I moved too fast or something? I think he thought I was going to do something …
L: How do you feel about that guy, and how do you feel about cops, after that experience?
C: My view on police officers is the same I view about lots of things. Fire is an important necessary thing in the world. But you must respect it, because it could burn you. When I see a cop, I see a giant ball of flame, and I try to avoid it.
L: Some people might see, a nice thing to roast marshmallows and family over. … Okay, I’m going to go dark; I apologize. Is it easy for this experience to turn to hatred?
C: Yeah, of course. And one might argue it’s an illogical conclusion. OTOH, you’ve got to live in the world. Hate is something that takes a lot of energy. One should reserve it for when it is useful, and not carry it around with you all the time. … The sad delusion .. that you can not worry about a car that is barreling toward you. We all have to be a little deluded, or else we’re paralyzed. But we shouldn’t be ridiculous. You said something earlier about empathy. What I would ask other people to get out of this story, is to recognize that it is real. People would ask me to empathize with the police officer. I would quote back statistics, that police officer isn’t even the top ten most dangerous jobs in the U.S. … Some of them die by suicide; that means there’s something, something’s going on with them. I’d be happy to be empathetic. But if we step back from what I feel, and what they feel, and you step up a level. The real problem is, we’ve built a big structure, where it’s easy for people to put themselves into different pots, different clusters, and forget that the people in different clusters are basically like them. … How do you fix Twitter? How do you fix Facebook? How do you fix racism? It’s structural … you’ve got to be able to have conversations in a way that’s relatively safe. The hard problem is setting up the structures in the first place. It’s in almost no one’s interest to change the infrastructure.
Racial tensions
L: I think leaders have the role of … uniting everybody with a vision; as opposed to divides in a vision. This particular moment in history feels like there’s a non-zero probability of a civil war. One of the … you may be able to speak to this better than me. One thing that’s not spoken about, I think, much, is the quiet economic pain of millions that’s growing, because of Covid, because of closed businesses. That’s building. Then, there seems to be an elevated level of emotion — from which the protests percolated. Why now? This moment in history?
C: Because time has passed. The first race riots were in Boston—Boston used to be the hotbed of riots, or so I’m told. … The point is, Basically, you’ve got to get another generation, old enough to be angry, but not so old to be able to remember what happened the last time. You said two things that I think are worth unpacking. One has to do with this moment in time. … The other has to do with the economic reality of Covid. I want to separate those things. This happened before Covid, right? … Let me preface all this by saying that although I am interested in history (it was one of my three minors as an undergrad—the other was Spanish; and the other was cognitive science) … Anyway. I’m interested in history, but I’m hardly a historian. Forgive my … I’ll ask the audience to forgive my simplification. I think the question that’s always worth asking; it’s the same question, but a little different. It’s not why now, but why not before? Why the 1950s, 1960s Civil Rights movement and not the 30s and 40s? Well, there was a civil rights movement before … these things had been building up forever.
L: It could have easily happened right after WWII …
C: I think the big difference was TV. These things were visible; people can see them! It’s hard to avoid. Why not James Farmer? Why MLK? Because one was born 20 years after the other. You know what King’s big failure was? It was in Georgia. They were trying to integrate; I forget the guy’s name. A sheriff made a deal with the whole state; we are going to take people, we’ll nonviolently put them in trucks, and we’ll take them to jails far away from here. … So, that didn’t work. But then they went to Birmingham Alabama, and people were hit by firehoses … So part of the delusion is pretending nothing’s happening. A large part of what happened right, was it was too public to ignore. … Part of the delusion was it wasn’t going to affect the West and the Northeast. … What’s happening now …
L: I guess you’re implying; it seemed to have percolated the most, a single video, George Floyd. It’s fascinating to think that, whatever the mechanisms that put injustice in front of our face … Those mechanisms are the mechanisms of change.
C: OTOH, Rodney King. I seem to be the only person who remembers this; a little before the Rodney King incident, there was a police officer saying things were really bad in Southern California. He had cameras follow him, and the first night he did it, some cops shoved his face through a glass window. … More people are willing to believe; and there’s more and more ways to get evidence: body cams, etc. I invite you to read what people wrote, about violence is not the answer, after Rodney King. It’s the same words over and over again. I’m surprised no one got flagged were plagiarism.
MLK vs Malcolm X
C: Did you know Malcolm X was older than MLK? He died sooner. But only by a few years. People think of MLK as the older statesman, but that’s more of a narrative device; not true at all. I reject the choice. I think it’s a false choice. … Violence has worked.
L: Yeah, that’s the annoying thing. It seems being irrationally upset has made progress.
C: It’s how you get people to notice you. Without getting people to notice you … maybe if TV didn’t exist, the Civil Rights movement doesn’t happen? Maybe if social media doesn’t exist, a whole host of things don’t happen? What do any of those things do except expose things to people. Violence is a way of shouting … Violence is the voice of the unheard, right? It’s a thing people do when they feel they have no other option. … Sometimes we agree; sometimes we disagree. Sometimes they’re justified, sometimes they’re not, we think. … Another way of putting it, that is less provocative, but I think is true, is that all change, particularly change that impacts power, requires struggle. The struggle doesn’t have to be violent.
L: The powerful don’t give up power easily.
C: Why should they? And this isn’t just about political change. This is true for understanding calculus. ..
L: We’re back to talking about faculty hiring.
Will human civilization destroy itself?
L: Do you think everything’s going to be okay.
C: Yes, I do. My mother says, and I find it quite comforting—this too shall pass. That doesn’t always mean this bad thing is going away. Good things, too. My 15-year-old daughter; I get to hang out with her now. … You learn to look back on something; mostly, it’s a memory. Yes, I think it’ll be okay.
L: What about humans? Do you think we’ll live into the 21st century? Are you worried we might destroy ourselves? With AGI?
C: I’m not worried about AGI doing it … I tend to think outside of things completely beyond our control, we have a better chance than not, of making it.
L: I talked to Alex Filipenka … about comets. … We’re watching comets; they could come out of nowhere, and all of a sudden, we have less than a month.
C: I’m choosing to believe that it’s going to be okay, and we’re not going to be hit by an asteroid.
Fear of death and the passing of time
#136: Dan Carlin, hard-core history!
Dan Carlin, host of Hardcore History and Common Sense podcasts. HH is one of, if not the greatest podcast ever made. Dan and Joe Rogan caused me to fall in love with the medium of podcasting … Meeting him was surreal. His voice has been a guide through some of the darkest moments of human history. … I think we’re living through one of the most challenging moments in American history. To me, the way out is through reason and love. Both require a deep understanding of human nature and human history. I am perhaps hopelessly optimistic about our future. … If we stand before the Great Filter, watching our world consumed by fire … think of this podcast is the final meal before the apocalypse.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Nature of evil
L: Do you think human beings are fundamentally good? Or are all of us capable of good and evil?
D: How do we define evil? Evil seems to be a situational eye-of-the-beholder kind of question? That could be a whole show, couldn’t it—defining evil.
L: That’s a slippery one. There’s some way your presence in the world leads to pain, suffering, and destruction from many others. You steal the resources and use them to cause more suffering than there was before. I suppose it’s related to this other slippery word, “suffering.”
D: I fully see where you’re going; here’s the problem: the question is the reason for inflicting suffering. Sometimes one might inflict suffering on one group, in order to maximize lack-of-suffering for another group. Or one might make the rational choice to inflict suffering on a small group to avoid that on the larger group.
L: That’s a thing; I’ve read Joseph Katkin, a Josef Stalin scholar. I’m not sure about Hitler, but with Stalin, it really seems that he was sane, and he thought he was doing good for the world. I really believe, from everything I’ve read about him, he believed communism was good for the world … …
D: I’m fascinated by the concept. Our very first HH, which lasted 15 minutes, was Alexander versus Hitler. The question was the motivation. If you go to a court of law because you killed someone, the question is, why did you kill them? In the show, we wondered. If you believe Hitler’s writings; it’s about as believable as any other political tract. … But the things he wanted to do, were in his mind, better for the German people. One of the rules, the reason why he did what he did, in other words to glorify himself. … … The motivations of the people doing these things, I think matter. I don’t think you can just sit there and go …
Is violence and force fundamental to human civilization?
Will we always have war?
The Russian front in World War II
Ideologies of the US, the Soviet Union, and China
Putin
Journalism is broken
Genghis Khan
Greatest leader in history
Could Hitler have been stopped?
D: … the times were so terrible; the options for operating within the system in a non-radical way were totally discredited. And then the alternatives to the Nazis … were Communist agitators. The average German person had three options. The discredited government put in power by your enemies. You could go with the Nazis, superpatriots calling for … Hitler was able to triangulate in this realm. He came off as a person who was going to restore German greatness. But if you don’t need German greatness restored doesn’t resonate. The reason your love idea I don’t think would have worked in the time period, is that’s not a commodity the average German was in search of then.
L: Interesting to think whether greatness could be restored without mechanisms that today .. are so evil.
D: But the … heartlands of Germany were occupied by the French. You weren’t able to have the hate dissipate; every time things were … the reparations were crushing. These things prevented … love from taking hold. Even if there were Germans that felt that way, it is/was hard to overcome the power of everyone else. Humanity—I believe on individual levels we’re capable of everything. But collectively it’s different, and messages of peace on earth and love your enemies … were overwhelmed and drowned out by the bitterness, the hatred—and the sense you were continually being abused by your former enemies. There were those on the Allied side that realized this; they said, we’re setting up the next war! … … I think love is always a difficult option. In the context of those times, it was more disempowered than normal.
L: Just to linger on it a little longer; the question of the inevitability of history … Do you think Hitler could have been stopped? There was a pain that was building, a hatred that was building. Was there a way to avert … I mean there’s two questions—could it have been worse?
D: The most logical thing—this brings a wonderful bowtie into the discussion. We’d talked about force and counterforce. The most obvious and much-discussed way he could have been stopped. When he re-militarized the Rheinland … what a couple of French divisions could have done if they had gone in and contested. If they had, Hitler was in no position to do anything about it. People who say; Churchill was one of these people. “He should have been stopped, militarily, when he was weak.” There were those in the … Weimar Republic that maybe could have done things. I do think had the French responded militarily, he would have been thwarted. I think he himself believed this would have led to his downfall. What I don’t like about this is it almost legitimizes military intervention at a very early stage. It should also be pointed out, there was a lot of sympathy on the part of allies that the Germans should probably have Germany back. The love and sense of justice on the Allies’ part may have stayed their hand. But, but, but, if the times were such that a Hitler resonated, simply removing Hitler would not have removed the context of the times. You could have had another one; or you could have wound up in a situation equally bad, in a different direction. It is hard to imagine anything could have been any worse. … Hitler’s an example people compare the Great Man theory of history versus the Trends & Forces theory of history. … If you take him out, but the trends are still there, does somebody else walk through that door …
L: The probability that charismatic leaders emerge … I’m torn on it.
D: The institutional stability of Germany in that time period was not enough to push back. If Hitler arose in 1913, he doesn’t get anywhere, because German’s institutional power is enough to quash that. … that would have prevented radicalism from getting out of hand.
Hitler's Antisemitism
L: WWII versus the Holocaust. We were talking about how history unrolls itself. I don’t know what to think about Hitler without the Holocaust. In his thinking, how essential the antisemitism and the hatred of Jews was. It feels to me that, we were just talking about, where did he pick up his hatred of the Jewish people? There’s stories in Vienna; picking up the idea of antisemitism as a really useful tool. As opposed to really believing it in his core. Do you think WWII and Hitler would be possible without antisemitism? Could we have avoided the Holocuast? Or was it an integral part of fascism and the Nazis.
D: Not an integral part of fascism; because Mussolini … That’s the big anomaly. Antisemitism didn’t need to be a part of this at all. Hitler had a conspiratorial view of the world. He thought the Jews controlled things. They were in charge of Bolshevism on one side, and capitalism on the other. The U.S. was Jew-ified. He had that line; if the Jews of Europe force another war in Germany … the Holocaust is a weird, weird sidebar. It weakened Germany. Look at the First World War; the Jews fought for Germany! Who was the most important Jewish figure that would have fought for the Germans if they hadn’t been antisemitic … Of course, to say any of these areas were not anitsemitic would be injustice to history… they had pogroms … It’s standard operating procedure; Hitler was a spike, ‘cause the government has a conspiracy theory. Hitler thought of them as weak and super-powerful at the same time. The whole idea of the blood and how it connects to Darwinism is just weird! Einstein, let’s just play with Einstein. If no anti-semitism in Germay, or none above the baseline level, does Einstein leave along with all the other Jewish scientists? And what does Germany have as increased tech. capacity if they stay? It’s a tragic fall in the Hitlerian worldview. You had mentioned earlier, maybe it wasn’t integral to his character but just a tool for power. I don’t think so. This guy became absolutely obsessed with this—with Jews. He surrounded himself with theorists who believed this too. You had a cabal of people reinforcing this idea that Jews control the world. He called it International Jewry. They were an enemy within. IT’s a nutty conspiracy theory, that the government—the big thing with Germany was culture. That they could be overtaken with this wild, wicked conspiracy theory … He took vast German resources and used them to wipe out this race, and he needed them for other things. That’s the weirdest part of the Nazi phenomenon.
L: Dark thing to think; maybe the Holocaust is what kept the Germans from getting nuclear weapons first.
D: If it weren’t so overlaid with tragedy … he sewed the seeds of his own destruction. The tragic flaw.
Destructive power of evil
L: And my hope is … a discussion I’ve had with my dad, a physicist. Is that evil inherently contains with it that kind of incompetence. My dad, an engineer, his belief is that, at this time in our history, the reason we haven’t had a nuclear terrorist is that the kind of people that would be terrorists are simply not competent enough at their job of being destructive. If you plot it, the more evil you are … If you were to consider the hatred of Jewish person’s evil … Just pure hatred of something that’s … grounded in conspiracy theories. If that’s evil, the more you give in to conspiracy theories, the less capable you are of deploying nuclear weapons. So that’s a hopeful message. Destructive people … are incompetent at creating ultimate destruction.
D: I don’t agree with that. … You say it’s been many decades. That’s like saying, in the life of a 150-year-old person, we’ve been doing well for a year. Bertrand Russell said it’s unreasonable to expect a man to walk on a tightrope for 50 years. The problem is, this is a long game. Until 30 years ago, the nuclear weapons in the world were really tightly controlled. … When we call these terrorists evil, it’s easy for an American to say Osama bin Laden evil. But one man’s terrorists is another man’s freedom fighter. What Osama did, we would call evil genius. The idea of hijacking planes; that he could pull that off. I’m still stunned by that. Here’s the funny part. I hesitate to talk about it, because I don’t want to give anyone ideas. You don’t need nuclear weapons to do incredibly grave amounts of danger. What one can of gasoline and a Bic lighter can do in the right time, can bring down societies. … The importance of the stability that a nation-state provides. When we went in and took out Hussein, one of the great arguments against it is he was a great anti-terror weapon in that region that you could have; they were a threat to him, so he took them out, using means much more repressive than we would ever be. This is the old line of why we supported right-wing death squad countries. … … The idea that the stability created by powerful and strong centralized leadership; it’s almost like outsourcing anti-terror activities. Same thing in Syria, with the Assads. You can’t have an ISIS in that area, because that’s taken care of by the Assad government. I would suggest the game is still on, whether these people get nuclear weapons in their hands. They don’t need them, I would suggest. If you think like the Joker in Batman; it’s scary to think how vulnerable we are.
L: The point is, your ability to generate good ideas, how you can disrupt a power grid, attack our psychology—with a can of gasoline, like you said .. coming up with good ideas there …
D: Are we saying evil people can’t come up with evil-genius ideas? I don’t think history backs that up. I mean, evil genius—that’s almost proverbial! I don’t want you to leave this in a terrible mood because I push back on every hopeful idea you have.
Will human civilization destroy itself?
L: That goes to the definition of evil. I believe in order to be good enough to be perceived as evil, you have to be able to construct an ideology, under which you’re doing good for the world. It’s difficult to construct an ideology … It’s like the question about aliens … if it’s possible to be evil, why haven’t we destroyed ourselves? You’re saying, the game is still on; the tools with which we could destroy ourselves are recent. … how soft we’ve gotten in terms of our deep dependence upon the system. Someone mentioned to me, what happens if the power goes out for a month? A Berkeley faculty member I was talking to who’s observing solar flares. It’s very possible a solar flare could knock out the power grid, for months. So, just as a thought experiment, what happens if power goes out for a week in this country? … Maybe that’s an act of nature, and even the act of nature will reveal .. the fragility of it all. And then the evil can emerge … Especially during a divisive time.
D: At baseline level, that would mean the entire supply line would start to break down … and then there’s a threat of everything.
L: If you were to bet money on the way that human civilization destroys itself, or it collapses in some way, the result would be unrecognizable to us—what would you say? Is it nuclear weapons? Is it more traditional times of war? Engineered pandemics? Nanotechnology? Artificial intelligence? Some sense of how humans might destroy ourselves?
D: I think what governs my view of this thing is our ability to focus ourselves collectively. I can ask, what are the odds that we’ll do X versus Y. Look at the 1962 Cuban missile crisis. We stared the potential of nuclear war right in the face. You want to talk about a hopeful moment. One of the rare times in our history where the odds were overwhelmingly that there would be a nuclear war. I’m not the Kennedy worshipper (that many are), but he by himself probably made decisions that saved 100 million or more lives. Everyone around him thought he should be taking the road that would’ve led to those deaths. To push back against that … When we talk about how the world will end; the fact that one person actually had that in their hands meant that it wasn’t a collective decision. Remember I said I trust people on an individual level; when we’re together we’re more like a herd. The higher ethics of an individual human being … When we have to act collectively, I get more pessimistic. Think about what we’re doing to the planet. We talk about it in terms of climate change. Forget that. Just look at the trash! We’re destroying the planet because we’re not taking care of it. Because to take care of it would require collective sacrifices. Too many people would have to be on board. Around the world! You can’t say we’ll stop doing it only if China does it! You get pessimistic hoping for those kinds of shifts. Unless—Krypton’s about to explode! If you’re talking about a gambling man’s view of this, that’s got to be the odds-on favorite. The systems aren’t even in place. The fact that we’d need intergovernmental bodies that are completely discredited now, on board. The amount of things that would have to go right in a short amount of time; you don’t have 600 years to figure this out.
Elon Musk, Tesla, SpaceX
L: Absolutely. I believe, like you said with individual charismatic leaders will save us. It starts with one leader, and their charisma inspires other leaders … Other spikes of leaders emerge. I tend to believe that, when you heat up the system, and shit starts getting really chaotic, whatever this intelligence we’ve developed—the leader will emerge.
D: Isn’t there as much of a chance that the leader will emerge and say, “The Jews are to blame.” I guess I’m saying, you could be right. But …
L: My intuition about the evolutionary process … the love in the system versus the hate in the system, the love is greater. The human kindness potential in the system is greater than the human hatred potential. In the time that it’s needed, the leader that inspires love … the Hitlers emerge, but in the grand scheme of history, they’re not that impactful. It’s weird to say, but not that many people died in World War II. Maybe 3, 5% of the population on earth, it’s not destructive to the entirety of human civilization. I believe that the charismatic leaders, when time is needed, that do good for the world, in the broader sense of good, are more likely to emerge.
D: It’s possible though … I think maybe you’ve divided this into too much of a black-and-white dichotomy, with love and good on one side. Let me throw in something that’s in the center of that dichotomy: self-interest. The good version we call enlightened; the bad version we call selfishness. I think simply a question of what’s good for me, or for my country, or for my point of view, or for my business. If I work in coal, and you tell me I have to stop doing that, because it’s hurting the earth; I have a hard time disentangling that question between the good of feeding my family … We’re not all going to decide at the same time that the interests that we have are aligned.
L: But I’ve looked at Ayn Rand; how bad can things go when everyone’s acting selfishly. We’re just two ants with microphones. … What do I mean by love and kindness? I think it’s human flourishing, on earth and throughout the cosmos. Whatever the engine that drives human beings is more likely to result in human flourishing. People like Hitler are not good for human flourishing. Maybe it’s intuition that kindness is an evolutionary advantage … It seems like, for us to multiply throughout the universe, it’s good for us to be kind to each other. .. Those leaders will emerge to save us from the Hitlers of the world.
D: You brought up evolution several times … I think going back to animal times, we are trained to deal with big threats right in front of us. If Krypton’s about to explode, I think humanity would be able to give power to the people that needed. But that’s what I think makes the pollution threat so insidious; it happens slowly, it defies flight-and-fight mechanisms. It requires an amount of foresight, that while some people would be capable of it, most others would be too worried about today’s threat or next generation’s threat. … Like you said, what’s the greatest threat. ……….
L: Elon Musk is a big fan of Hardcore History
D: I know Elon
L: Awesome; that’s relationship goals; listen to HH on the weekend with your loved one. If I were to look at (Elon) from the perspective of human history, seems like he’s one that will be remembered. Who are the people we’ll remember, whether the Hitlers or the Einsteins. If I was to guess what Elon will be remembered for, I think maybe the work he’s doing with SpaceX; as the person who launched a new era of space exploration. If we’re able to venture out into the stars. Do you think humans will be a multiplanetary species? Do you think he’ll be successful in his dream? He really wants us to colonize Mars. And then, other earth-like planets throughout the galaxy.
D: It’s hard to get your mind around (SpaceX), because he’s doing what it used to be governments doing before. Pushing the envelope faster than the governments at the time were moving. A lot of people think Elon is overrated, but you have no idea. But that’s actually not what I’m most impressed with. It’s Tesla I’m most impressed with. We talked about the environmental stuff, and our inability, all at the same time, to sacrifice our self-interest for the goal. In my mind, what he’s done is recognize that problem, and instead of building a car that’s a peace of crap, but it’s good for the environment—he’s trying to create a car that, if you’re only motivated by your self-interest, you’ll buy it anyway! … One of the things this pandemic has done, is show us how amazingly quickly the Earth can rejuvenate itself. So what if, to name just major pollution source, we didn’t have the pollution caused by automobiles. … What if you created a vehicle that’s superior in every way? That’s the best way to get around that problem of people not wanting to sacrifice. As he’s told me before, you know the last time a car company was created that bla bla bla? He’s right! Even though he’s pushing the envelope on the space thing, I think someone else might have done that some day. But I’m not sure any of these car companies would have gone where he’s leading them … who was going to do that? I hope he doesn’t hate me for saying this; I think the Tesla idea may alleviate the need to get off this planet. The Tesla idea, I think, is the best chance that we’ll be around long enough to see Mars colonize.
L: Another thing from my perspective as someone that’s starting a company. He serves as a beacon of hope … to push back on our doom conversation from earlier. That a single individual could build something, that allows us as self-interested individuals to gather together in a collective way … it gives me hope as an indvidual, that I could build something; that I could have impact that counteracts the Stalins, the Hitlers, all the threats that human civilization faces. I didn’t believe the individual has that power in the halls of government. I don’t think one presidential candidate can unite the world. It feels like everything I’ve seen with Tesla, it can bring the world together to do good. That’s a powerful mechanism; whatever you say about capitalism; you can build companies … it starts with a single individual; there’s a collective that builds around that. The ___ of a single individual can catalyze something that takes over the world.
D: I think the genius of the idea is that it doesn’t require us to go head-to-head with human nature. He’s built it in; he’s saying, I’m not requiring you to be an environmental activist. … That takes into account our foibles as a species. That’s something that does turn off my doom-caster cynicism. You can find somebody who doesn’t even believe the environment’s a problem, and they want a Tesla.
Steering around the iceberg - wow do we avoid collapse of society?
L: A few voices of empathy can help us avoid … not somebody starting a fire and so on, but avoid the escalation of it; the preparedness of the populace to escalate those events; to turn a single event, or a riot, or a shooting—into something that creates ripples that grow, as opposed to ripples that fade away. So, being cognizant of the fact that a lot of very destructive things might happen in November. And a few voices could save us. By being the voice of calm, that like calms the seas from, or whatever the analogy is, from boiling up. ‘Cause I truly am worried; I somehow have felt that the American project would go on forever. When I came to this country, I have a dream of creating a company that will do a lot of good for the world. I thought America is the beacon of hope for the world; the idea of empowering companies that could do some good for the world. I’m worried about this America. Our family came from nothing. To be able to do anything in this country; I’m just worried about it. I feel like a few people could keep this project going. People like Elon. People like Joe. Do you have a bit of that hope?
D: I’m watching this experiment with social media right now. I feel like we’re all guinea pigs … I have two kids; one’s 18; the other’s 15. When I was a kid, people that were 18 and 15 would not be that different. You would classify them as being in the same generation. Because of the speed of technological change, there is a vast difference between them: the apps they use, their social skills. This is uncharted territory; we’ve never been here. As it relates to your question: reading how people treat each other on line. Some times it’s just trolling; I’m trying to figure out how, if this is how we have always been as people. We’ve always been this way, but we’ve never had the means to post our feelings publicly about it. Or .. if the media has provided a change, and changed us into something else. Either way, when one reads how we treat each other, and the horrible things we say about one another on line; they have a cumulative effect. I was reading Megan Markel; she was talking about the abuse she took online, and how incredibly overwhelming it was, and how many people were doing it. Okay, this is something people … never used to have to deal with. This is the ultimate doom-caster thing to say. When you think of historical figures pushing love and peace, and pushing bridges between enemies. … What happens to Gandhi, and Jesus, and Martin Luther King? You think about all those people? It’s ironic that people that push for peaceful solutions are so often killed? But it’s because they’re effective. And why are they killed? Because their effectiveness—they’re so charismatic, if you eliminate them from the scene, the odds are you won’t have another one for a while. .. The odds are they will destabilize systems; those systems have a way of protecting themselves. See history’s pretty pessimistic, I think, by and large. People who are dangerous to the way things are—tend to be removed.
L: Yes. I feel like you’re right. The ripples that love … in history, are less obvious to detect, but are more transformational.
D: One could make a case: the long-term ripples of Gandhi, they’re still affecting people today, I agree.
L: You feel the ripples … But, even if that’s not true, I tend to believe—and by the way, the company that I’m working on, is exactly attacking this—is a competitor to Twitter. I could build a better Twitter.
D: A three-year-old child could build a better Twitter. Facebook, too. They’re really awful platforms for intellectual discussion, meaningful discussion … And I’m on it.
L: I tend to believe that we live in a time, that the tools, that people who are interested in providing love—the weapons of love are much more powerful. The one nice thing about technology—it allows anybody to build a company that’s more powerful than any government. I tend to believe that somebody like Elon, somebody like me, could have more power (to affect change) than any one government.
D: How do you deal with the fact that already governments who are afraid of this are walling off their own Internet systems as a way to create firewalls, to prevent you from doing what you’re talking about doing. There’s an old line that if voting really changed anything, they’d never allow it. If love, through a modern-day successor to Twitter, would really do what you want it to do, and this would destabilize government, do you think governments would take counter-measures?
L: To push back on something you said earlier, I don’t think love is as much an enemy of the state as one would think.
D: Different states have different views.
L: I think the states want power; I don’t always think love is in tension with power. I think it’s not just about love; it’s about rationality, reason, empathy; all of those things … don’t have to be in conflict with each other. That’s one sense, I feel like—you could Trojan horse love into. But you have to be good at it. You have to be conscious of the way these states think. … The fact that China banned certain services … that means the companies weren’t eloquent … weren’t actually good at infiltrating. Isn’t that a song? Love is a battle field. It’s all a game, and you have to be good at the game. Just like Elon, with Tesla and saving the environment. It’s not just by getting on a stage and saying, it’s important to save the environment. It’s building a project that people can’t help but love. There’s a game to be played.
D: OK, let me tell you a story of the 1960s … a revival of neoromantic ideas. I have a buddy who went to the protests … I was romanticizing it; he said don’t. Most of the people who went to those protests; all the people there were there to meet girls and have a good time. But, it became in vogue. Let’s talk about your empathy and love. You’re never going to grab that great mass of people who are only in it for themselves. But if meeting girls for a young teenage guy requires you to feign empathy; requires you to read deeper subjects because that’s what people are into. You could make empathy trendy, love trendy. The possibility exists to change the zeitgeist, and reorient it in a way … Does that make sense? So we’ve found a meeting of the minds.
L: Yes. Creating incentives … that encourage … It all boils down to meeting girls and boys.
D: Once again, you’re getting to the bottom of the evolutionary motivations … It’s difficult …
Advice on podcasting
L: There’s something about hardcore history. I do some crazy challenges; running and stuff. I remember one of my favorite episodes, the Painful-tainment one … I remember listening to that when I ran 22 miles; for me, that was a long distance. It just pulls you in; there’s something so powerful, about this particular creation that’s bigger than you, that you’ve created …
D: Anything that is successful like that becomes bigger than you; that’s what you’re hoping for, right?
L: A question I have. If you look at a mirror; but you also look at me … What advice would you give to yourself? To me? To other podcasters—about this journey that we’re on. It feels like podcasting is special; what advice do you have, for people that are carrying this flame.
D: I’m often asked for advice by new podcasters; I have a tried and true list. But I don’t have advice for you, or for Joe. Joe’s figured it out.
L: He’s still a confused kid …
D: But that’s the genius of it! That’s what makes it work. By the time you’ve reached the stage you’re at, or Joe’s at—you don’t need advice. .. You found your niche; you don’t need me to tell you what to do. I might ask you questions about how you do what you do, right?
L: We were talking offline about monetization … One of the things Joe is facing, I don’t know if you’re paying attention; he joined Spotify with a $100 million deal … I don’t give a damn about money, personally, but I’m single; and I like living in a shitty place.
D: You get the freedom to not care! Not saving for anybody’s college …
L: Okay, so maybe it’s romanticization, but … when I heard about Spotify teaming up with Joe, I thought, “fuck the man!” But then I thought, you can’t lock up this special thing we have. And maybe, these are vehicles for reaching more people, and respecting podcasters more. That’s what I mean by, it’s unclear what the journey is. Now there’s one million plus podcasters. I wonder what the journey is. … Do you have a sense; are your romantic in the same kind of way? You have roots in radio, too. Do you feel podcasting is pirate radio? Or is the Spotify thing one possible avenue … Are you nervous about this for Joe? Or you think it’s a good thing
D: My story: I was in radio; I started a company when the dot-com boom was happening. I did it with six other people; the whole goal of the company; what we were pitching to investors was something called amateur content. My job was to be the evangelist. I would go to these people and sing the praises of all the ways amateur content was going to be great. I never got a bite. They told me, anybody who’s good is already going to be making money with this. I said, forget that, we’re talking about scale here. But the podcast grew out of that, because if you’re talking about amateur content in 1999, you’re ahead of the game in seeing where it’s going to go technologically. WE started a political podcast in 2005. We didn’t have any financial ideas; we were just trying to get a handle on the technology … how do you get enough money to support us while we’re doing this. … Once we started down that, we figured out other models. We saw the old shows. All of this became ways for us to support ourselves. As channels and tech developed … it became easier to podcast and build an audience, but it became more difficult to make money off of it. There’s a lot of people doing similar things. We used to just sell MP3 file. Now, every time there’s an OS upgrade, something breaks for us. So, I may want to start hiring staff, more staff …. IT’s becoming hard to do it lean and mean. If somebody like a Spotify comes in and says, “Hey, we’ll handle that stuff for you.” In the past I would have said eff off, we don’t need you. But it’s becoming onerous for me … if someone were to come in and say, we’ll pick that up for you; we won’t interfere with your content at all … everybody’s design is different, but as a longtime pirate podcaster, we’ve been looking to partner with people …
Joe Rogan, Spotify, and the future of podcasting
D: I’m always looking for ways to take some of this off my plate. We’re trying to do art; there’s something satisfying in that. But I can’t stand the increasing amount of time the monetization takes … If partnering with outside firm works without interfering with your ability to do the art … it’s enticing! I don’t like big companies, so I’m afraid of whatever strings may come with that. If I’m Joe Rogan, talking about (touchy subjects) I’d be nervous … but people don’t understand … all the operating systems, the new pod catchers that make it easier for people to get the podcast … I think maybe the short answer is, as the medium develops, it’s becoming something you have to consider. Not because you want to sell out, but because you want to keep going.
L: The thing that convinced me … if you walk into this whole thing with skepticism of big companies, then it works. Spotify appears to understand the magic that makes podcasting; at least they understand enough to respect Joe Rogan. … There’s the Internet, and there’s people on the Internet, and they have opinions about Joe. But there’s two important things. One, Spotify literally doesn’t tell Joe anything …
D: Didn’t he insist upon that?
L: Also, companies have a way … “we’re not forcing you,” …
D: I hate that!
L: Spotify’s smart enough not to send a single email of that kind. There are meetings inside Spotify where people complain, but those meetings never reach Joe. The idea that Spotify is different than pirate radio … nobody gives a damn about this; the nice thing about Spotify is, they want Joe’s podcast to succeed anymore. That’s the difference between YouTube and Spotify. Spotify wants to be the Netflix of podcasting. What Netflix does, they don’t want to control you in anyway ..
D: But their interests are aligned. .. You brought up YouTube and Spotify … To me, YouTube … I’ve been around podcasting so long now, that I’ve seen rounds of consolidation over the years. I’m not going to mention any names … until recently, the consolidation was happening … By deciding to consolidate your materials in a walled garden … Your choice is, I’m going to accept money with this company … but then I lose some of my audience size. But … when you get to the level of a Spotify, these are people that can potentially, potentially enhance your audience over time. So the risk is low to you; if you decide after a year or two, that you’re done with them and you want to leave … As opposed to with some of the other of these walled gardens .. you may in two years have a larger audience. It takes away some of the downside risk …
Future episodes of Hardcore History podcast
L: I don’t love asking this question, but it’s probably the most popular, unanswerable question … YOu said you don’t release shows very often; the request, and the question is, “Can you tell Dan to do one on the Civil War? ON Napoleon Bonaparte.” You’ve spoken to this. You said, re: the Civil War, that as a military historian .. you like when there are contrasts between the sides. In that war, there’s not an interesting contrast. You also said the shows you work on, you have some roots of fundamental understanding about that period; when you work on a show, it’s pulling at those strings further …
D: Wow, you’ve done the research; these are words out of my mouth.
L: Are there ideas, strings in your head, that may be possible future episodes? Not committing to anything, are there any loose threads that you might want to pull on?
D: Yeah; there are things we keep in our back pocket forever. The Blueprint for Armageddon … we had that in the back pocket for a while.
L: That was a great series
D: Kicked my rear end, man. I think it’s 26 hours … and we don’t do scripts; it’s improvised. Some people acting annoyed that we don’t have enough episodes … but the last show we did about the fall of the Roman Republic was 5.5 hours! That’s a book. And that was Part VI!
L: On World War I, it felt like you were emotionally pulled into it.
D: That was the feeling that the people at the time felt. … That had sort of inadvertantly; when you improvise a show … it had created the right climate to create empathy with the story line … That, to me … wouldn’t have happened if we scripted it out.
L: So you bring the tools of knowledge to the table, and then in large part improvise the actual wording.
D: I do have notes about things: “On p. 427 of this book.” Sometimes I’ll have notes on, here’s where I left off yesterday. The improv allows us to go off on tangents. I’ll try things. I’ll spend two days going down that road; and then decide it doesn’t work. People questioning our dedication have no idea how many things we’re throwing out! … There goes six weeks of work! … Here’s the thing; our show’s a little different than other people’s; Joe called it Evergreen content. These shows are just as good (or bad) five years from now as they are now. … When you’re creating Evergreen, you have two audiences. The one that’s waiting for the next show. But the show’s also for people five years from now … who don’t care a wit … they care about quality. The audience will forgive me if it takes too long but is really good when you get it. They won’t forgive me, if I rush it to get it out on time, but it’s a piece of crap … My job is quality; everything else goes by the wayside. That’s what gives you longevity …
D: … we had a chapter in that book called “Pandemic Prologue.” And I had a guy say to me, did you ever think how lucky you got on that? If I’d had my druthers, I would have taken longer writing that book; and then that chapter wouldn’t have looked good. But to me, it didn’t look the way I wanted it to look at yet.
L: Can you drop a hint of things you’re keeping on the shelf
D: Oh the Alexander the Great thing. … I’ve talked around the career; I’ve talked about his mother in one episode. I’ve specifically left this Alexandrian sized hole in the middle. I’m going to lovingly enjoy talking about this crazily interesting figure … When we do this World War II: Asia and the Pacific; when we finish this, the tendency is to move to a much different period because we’ve had it and the audience has had it.
L: What about the other half of that .. episode. You’ve done quite a bit about the war, and Germany. Would you ever think about doing Hitler, the man?
D: One of my standards for not getting five-hour podcasts done, was to flit around the interesting points. We didn’t realize we were going to get an audience that wanted the actual history … we thought we’d be focusing on the weird stuff. We did a show called Nazi Tidbits. … We dealt a little with it. It would be interesting. I’ll give you another example; history is not static. We had talked about Stalin earlier. Our portrayal of Stalin had been … outdated. When you talk about Hitler ,it’s very interesting to think about how I’d do a Hitler show today versus how I would have 10 years ago. There’s lots of new stuff, lots of new scholarship.
L: I have an unusual amount of power, because I have you stuck in a room. After Alexander the Great, I hope you’ll do a return to Hitler. The Rise and Fall of the Third Reich, I really love that study of the man of Hitler. Perhaps an episode that’s really focused on a particular period. It’s funny; he’s one of the most studied period, but I feel like all the stories, or most of the stories haven’t been told.
D: I’ve been reading a lot about Hitler. When I read that book you’re reading; I thought I needed to go to a psychologist. … You would think something like that is pretty established fact, but there’s new stuff coming out; Germany’s been investigating this guy forever. I took five years of German in school; I can’t read any of it. when you talk about fascinating figures, the whole thing is twisted-ly weird. …
Is Ben real?
L: Who’s Ben, and is he real?
D: I don’t know what you’re talking about. That’s like asking me if Harvey the white rabbit is real.
L: Well, a lot of people demanded I find out a way to prove the existence …
D: If I said he was real, people would say no he’s not; and vice versa.
Meaning of life
L: What’s the meaning of it all? You study human history … Have you been able to make sense of why we’re here on this spinning rock?
D: What I find interesting is certain consistencies we have over time. History doesn’t repeat but it has a constant, and the constant is us. Now, we change. … Social media, that’s a new element. We’re still motivated by love, hate, greed, anger, sex … That which connects us with the ancient …. The human element; you mix it with systems … the ancient Roman republic, people point to all the time. You have humans, just like you had then, and then you have a system that resembles the one that we have here. You begin to see things that look like they rhyme a little. When you show us, 500 years ago in Asia, and 800 years ago in Africa. You look at all these different places where you put the guinea pig in … I feel like it helps me flush out a little bit more who we are in the long timeline … Who we’ve always been.
L: Do you think in that common humanity thoughout history, is there a why? underneath it all.
D: Now, you’re into the simulation thing. The idea that there’s some kid, and we’re the equivalent of an alien’s ant farm. I think the wise elude us … I’m not wise enough to propose a theory myself, but I’m interested enough to read all the other ones out there.
#137: Alex Filippenko, UCB astrophysics
An astrophysicist and professor of astronomy from Berkeley. He was a member of the Supernova Cosmology Project, and also a search team that used supernova to posit the universe was accelerating, and this led to the discovery of dark energy, for which he won a 2011 Nobel Prize.
As we talk in this conversation about the objects that populate the universe, they are both electrifying and terrifying. IN the chaos, tension, conflict, and social division of 2020, it’s difficult for us to forget how difficult it is to just have a conversation.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Universe expansion
L: Will the universe expand forever, or collapse on itself. We …
Dark energy
Scientific revolutions
Asteroid hitting Earth
Giant solar flares and the power grid
Elon Musk and space exploration
Exoplanets
A: … I just don’t see it happening. What I see potentially happening, if we avoid self-destruction, is that machines will do it—robots. They can hibernate for a long, long time as they’re traveling. If some energetic charged particle hits the circuitry, it fixes itself. It’s a form of AI: you tell the thing, fix yourself. When you land on the planet, start producing copies of yourself. This is futuristic; it’s more feasible, I think, than sending flesh and blood over interstellar distances, a quarter million years to even the nearest stars; you’re subject to all kinds of … radiation. That’s one of the problems of going to Mars; it’s not a three-day journey like going to the moon. You’re out there for the better part of a year, or two, and you’re exposed to a lot of radiation.
L: The hope is the AI systems might be able to carry the fire of consciousness; whatever makes us humans. … A little drop of whatever makes us so special. Not to be too poetic about it.
A: Yeah … is there something beyond the bits, the 1’s and 0’s, for us. I like to think there isn’t anything; that all our compassion … comes just from these 1’s and 0’s. To me, that’s a beautiful thought; the idea these silicon-based thing, our creations, could be our descendants. To me, that’s a beautiful thought; but others find it horrific.
L: Yeah. I believe it’s beautiful as well. … Whatever systems we create that take over the world; it’s impossible for me to believe that they wouldn’t create some aspect of what make humans beautiful. .. That naturally, the systems that will out-compete us on this Earth will be cold, not conscious, not capable of empathy, love, hate … the beautiful mix of what makes us human. To me, intelligence requires all of that. In order to outcompete humans, you’d better be good at the full picture. …
A: … able to survive the onslaught of existential threats … that we bring on ourselves, or that we don’t anticipate. … Really, if we want to move to other planets outside our solar system, realistically, that’s a better option than believing that humans will make these journeys. … You’d have to go close to the speed of light. Looking at E = mc^2 … the energy required to get us to that speed; that amount of energy is unfathomable. We can do it with protons at the large hadron collider. But that’s just a proton; we’re gazillions of protons. And that doesn’t count the rocket that would carry us, the payload. …
Traveling close to the speed of light
A: Zip past it, take a picture of the exoplanets that we know, and then send the images back to us. That’s a tiny little thing; maybe you can accelerate that, to, they’re hoping 20% of the speed of light. … That’s a very forward-looking thought; I support the idea. There’s a big difference between sending a tiny camera, and sending a payload with equipment … that could mine resources on the exoplanet that they reach …
Traveling faster than the speed of light
A: Wormholes, sci-fi … is that interesting to you? Is that a dream for you, that we might be able to find clues … I’m excited by the potential physics that suggests this faster-than-light travel; cutting the distance to make it short through a wormhole. Call me not-very-imaginative, but based on today’s knowledge of physics. I know people have gone down a rabbit-hole; I know Lord Kelvin said a century ago that all physics is done. Let me not be another Lord Kelvin. But to me, most of these schemes, if not all of them, feel … impossible. A non-rotating blackhole is a no-go; the singularity is a point-like singularity, and you have to reach it, and you get squished by it. Now a black hole … maybe you could traverse the donut-hole or ring singularity. So there’s two theoretical ways you could get through a rotating black hole or a charged black hole. Rotating black hole is definitely a reality. Do they have traversable wormholes? Probably not; when you go in, you go in with so much energy that it either squeezes the wormhole shut, or you encounter so much energy (due to mass inflation instability) that it vaporizes you. Now, maybe you make it through in vapor form; it’s still information, but it’s scrambled information. But then what bothers me; as soon as you have the possibility of traversing the wormhole, is that you could come back to your universe at a time prior to your leaving, and you could then prevent your grandparents from ever meeting, and then you’d never have been born—It’s a violation of causality, of cause and effect. I take causality violation very seriously.
L: You took a stand! … It’s not just philosophy
A; Back to the Future like … you come back to the universe in a way that does not affect your journey. Some say you end up in a different universe; this goes into the different … multiverse hypotheses. Then it’s not the universe from which you left. You’re not really going back in time to the same universe; you’re not going back in time or forward in time to the same universe … Then there’s this idea … of an air drive; you warp space time in front of you. The problem I see with that … aside from … energy requirements. Miguel Au-cubier (for whom it was named) acknowledged all this. Problem I see; you’d have to get to those places in front of you … But you’d have to get there in a normal way … In a sense, you haven’t saved any time …
L: Yeah, we’re assuming we can fix all the biological stuff … Ultimately, it might boil down to extending the life of the human in some form … or figuring out, genetically, how to live forever. That long journey might be different if our understandings were different
A: If you could put us in suspended animation … then you can do it; it’s still not easy; you’ve got some big old huge colony; through E = mc^2; that’s a lot of mass to acccelerate; the Newtonian kinetic energy … but at least you’re not trying to do it in a short amount of clock time … The energy is your rest mass, divided by the root of M^2 over c^2 … As v/c approaches 1 … then the other variable approaches infinity. If you wanted to do it in zero time, you’d need an infinite amount of energy …
Intelligent life in the universe
L: One of the thing that inspires a lot of folks, including myself, is that this conversation is happening on another planet, with intelligent life forms. First we could start—as a cosmologist; what’s your intuition about whether there is or isn’t intelligent life out there
A: I’d say I’m one of the pessimists. The observable universe goes out 14 billion years in light travel time; 46 billion years if you take into account the expansion of space. That encompasses 100 billion to a trillion galaxies … There may well be other intelligent life
L; But your sense is, our galaxy’s not teaming with life
A: Yeah … I’ll get to the primitive life, the bacteria, in a moment. We may well be the only ones in our Milky Way galaxy; I’d side with that school of thought. I don’t see human intelligence as being a natural evolutionary path for life. There’s a number of arguments. There’s been more than 10 billion species of life on Earth; nothing has approached our level of intelligence … whales and dolphins are reasonably intelligent, but they can’t achieve abstract thought, or build machines to study the world. We came about as early hominids … and as homo sapiens a quarter of a million years ago. .. For most of our history, an intelligent alien would pass by and not identify intelligence.
Fermi Paradox
A: That brings me to the fourth issue, the Fermi Paradox: If they’re common, where the hell are they? The various UFO reports don’t clear the bar of scientific evidence in my opinion … If the galaxy is teeming life, then you’d expect some of it to be as intelligent or more intelligent than ours … There could be civilizations that are billions of years ahead of ours … There’s no telling what they could have achieved in thousands and millions of years. … IF they reach this ability of interstellar travel, you can show that you could populate the whole galaxy; then you wouldn’t have had to try to go far to detect them
Finding alien life would be bad news
L: Exciting … finding life elsewhere like Europa
A: Yeah—that would be bad news; If we find lots of pretty-advanced life. Especially if it was fossilized … If we found intelligent or even trilobytes elsewhere, that would mean that the “great filter” is ahead of us. That lots of things had gotten roughly to our level … but given the Fermi Paradox; if you accept that it means that no one else is out there, and yet we’ve found lots of things roughly at our level … that bodes poorly for our long-term future.
L: I’m not sure if us being alone … is pessmistic
A: It’s good news, in a sense, because it means that we made it through
UFO sightings
L: One of the ideas that I find compelling … is they (UFO sitings) started showing up, around when we developed the capability to create nuclear weapons. There is something in the human condition that wants to see, wants to believe, beautiful things …. Bigfoot is a big fascination for folks. UFO sitings is an example of that; people look at lights in the night sky. It’s kind of a downer to think in a skeptical sense—that’s just a light. I felt that first when my dad, a physicist, first told me about Ball lightning. That fire, that excitement; … what do we do with that? Most of the scientific community rolls their eyes and dismisses it. Is it possible a tiny percent of those people saw something we need to investigate it.
A: Sure; we should investigate it. But they haven’t brought us evidence.
L: But David Fraber! … the conspiracy theory folks say, whenever they have good hard evidence, that the U.S. government will hide the good evidence. … Of course you want to protect military secrets; but I don’t know what to do with this beautiful mess. I think millions of people are inspired by UFOs
A: I would say as Carl Sagan used to say, extraordinary claims require extraordinary evidence. We would welcome such evidence. OTOH, a lot of the things that are seen, or perhaps even hidden from us for military/surveillance purposes. Maybe some of these pilots saw Soviet or Isreali satellites. Some crashes that occurred were later found to be weather balloons. When there are more conventional explanations, science tends to stay away from the sensational ones. It may be that someone else’s calling to study these things. I don’t categorically deny the possibility that ships of some sort could have visited us. … Our Voyager and Pioneer spacecraft are going to be in the vicinity of different stars. There’s nothing that breaks any laws … studying it slowly. But that’s different than actual aliens, altering the course of their vehicle; in real-time spying on us. They are likely many light-years away; they’re not going to have broken that barrier as well. Go ahead—study them! For some young kid, it might be their calling, and that’s how they might find meaning in their lives. I chose not to, because I found the evidence, at a very young age, quite unconvincing.
L: I also believe the U.S. government, especially this year, is not sufficiently competent to conceal all evidence of aliens being here
A: I meant originally that I’m pessimistic about the possibility that there’s many, many of us out there. It may mean in a good thing for our ultimate survival … But anyway, I think UFO research is interesting. I guess one of the reasons I haven’t been convinced is that there are some scientists investigating this and they haven’t found clear evidence. I can’t say for sure there’s a critical mass of such scientists. You never get these reports from hard-core scientists. Astronomists, we spend our time studying the heavens; you’d think we’d be the ones finding that in the sky, and we never do.
L: I try to keep an open mind; it’s actually really difficult for a scientist. This year, I’ve probably gotten over a thousand emails on the topic of AGI; it’s very difficult. People write to me, “How can you ignore this model—it’s obviously going to achieve general intelligence.” The problem is, it’s very difficult to wade through a bunch of b.s. It’s very possible that you actually saw the UFO; but you have to acknowledge the hundreds of thousands of people who are a little bit, if not a lot, full of b.s. From a scientist’s perspective, it’s a lot of hard work. Why investigate Bigfoot, when so much else is available?
A: In essence, there’s a million decoys. We get lots of lots of phone calls when Venus happens to be close to the crescent moon; we get phone calls: “It’s a UFO!” I’m not saying the best UFO reports are of that nature; no, there’s more convincing cases. But there’s so many decoys, so much noise that you have to filter out. There’s only so much time, and you have to choose what problems you work on.
Universe expansion speed
L: The radius of the observable universe is 47 billion light years. And the age of the universe is 13.7 billion years. That’s less than the radius of the universe. How is that possible?
A: Great question. … I meant to bring a ping pong ball and a rubber band. You have the space itself between clusters of galaxies expanding. Imagine light going from one cluster to another. IT traverses some distance; while it’s doing so, that part it already traveled through continues to expand. Now, 13.7 billion years might have gone by since the light we’re seeing from the early stages … the afterglow of the big Bang. … That’s how long it’s taken that light to reach us. But while it’s been traveling that distance, the parts that it already traveled continue to expand. IT’s like walking on a moving-walkway at an airport, but the walkway is continuing as well. You end up traveling the same distance faster … You get a roughly a factor of pi; it’s more like 3.2; you multiply the number of years the universe has been in existence by 3.25 or so, and that’s how you get this 46 billion light year radius.
L: How is that not traveling faster than the speed of light?
A: Because locally, at any point, if you were to measure the photon zipping past, it would not exceed the speed of light; it’s a locally-measured quantity. Yes, zoomed out, it looks like it traveled a greater distance. But locally … IF you give the light credit for having traversed that distance, then it looks like it’s going faster than the speed of light
L: That’s not how speed works (laughing)
A; On relativity; if you take two ping pong balls sufficiently far apart; you can have them moving away from each other, faster than the speed of light … …. That light limit … applies to a particle moving through pre-existing space. …
The universe is infinite
L: Could we talk about the Big Bang a little bit? The entirety of it. The universe. Was very small.
A; IT was not a point. If we live in a closed universe now—a sphere, or a hypersphere, then regardless of how far back in time you go, it was always that topological shape. You can’t turn a point into a shell. IT was always a shell. So to say it started as a point; that’s being flippant. We don’t know if it was finite or infinite; … if it exponentiated, and continued to do so, then it could in fact be infinite right now; most cosmologists think that it is.
L: Infinite in what dimension?
A: If I was trying to use light to measure its size; I’d never be able to, because it would always be bigger than the distance light could travel
L: But if it’s a sphere, how can that thing be infinite?
A: It expands exponentially; that’s what inflation theory is all about. … As explained by ___ ___ from MIT and ___ ___ from Stanford. IN an exponentially expanding universe, you try to make this measurement; the amount of space remaining to be traversed is always a bigger and bigger quantity, so you’ll never get there. So operationally, it can be regarded as infinite
L: That’s one of the best definitions, physical manifestations, of infinity
A: Yeah, how would you measure it? If I took a god-like slice in time, then wouldn’t it be finite. They object: you can’t go outside and take a god-like slice of time. What slice of time you’re taking depends on your motion; that’s true even in special relativity. Slices of time get tilted if you’re moving quickly; the axes actually become tilted; not perpendicular to one another. Look at Brian Greene’s lectures; he imagines taking a loaf of bread and slicing it in units of time as you move forward … IT’s not even clear what slices of time means. I’m an observational astronomer; the way I understand infinity is, operationally, there’d be no way of seeing that it’s a finite universe. In that sense it’s infinite; even if it started out as a finite little hypersphere.
What happened before the Big Bang?
L: But … it didn’t really start out there; what happened before that? Let’s go there!
A: WE don’t know.. .. whether there are other universes out there. I like to say these are on the boundaries of science. … We don’t wake up at 3 a.m. and think about it. We have real test-able physics we can use to draw certain conclusions that are plausibility arguments based on what we know. Admittedly, there are not really direct tests of these hypotheses. These aren’t elevated to theory; because a theory is something with a lot of observational support behind it … … They may have indirect tests, in that if you adopt this hypthesis; there may be a bunch of things you expect of the universe… .But we’re not really measuring anything at T < 0. … It’s hard to prove uniqueness; it’s hard to completely convince oneself … Let’s say there are 50 predictions one of these hypotheses makes … Let’s say 49 of them are successful; but you can’t get the 50th that says there are other observable universes. … That’s science. They might say, “Oh, you cosmologists—that’s not really science.” It’s not really testable, but it is sort of testable. It’s not that we’re coming up with crazy ideas … You might say we created a lot of energy … In fact, this quantum fluctuation out of nothing, in a quantum way violates conservation of energy, but who cares. An inflating universe maintains whatever energy it had. In essence, the stuff of the universe has a positive energy, but there’s a negative gravitational energy associated with it. … .. . it’s also gaining negative units of potential energy; the total energy remains the same. .. Other physicists say energy isn’t conserved in general relativity; another way out of creating a universe out of nothing. This is all based on reasonably well-tested physics. These extrapolations seem outrageous, but they’re within the realm of what we call science already. Maybe some young whipper-snapper will find a way to directly test what happened before T = 0
Roger Penrose
L: He has an idea that we might have information that travels from before the Big Bang. Do you think it’s experimentally possible to detect some radiation …
A: The cosmic microwave background radiation, there might be ways of doing that.
L: Philosophically or practically possible to detect signs it’s from before the Big Bang? Or … no.
A: It’s very difficult to answer right now; we don’t have a single verified, fully self-consistent, experimentally-tested, quantum theory of gravity … You need both quantum mechanics and general relativity. Same thing if our universe re-collapses … into another Big Bang. IDK if we can answer that right now; we don’t have a quantum theory of gravity that most physicists trust. There are various forms of string theory; there’s quantum loop gravity. … which will be the one that stands the test of time, of experiment and observation. My feeling is, these things don’t survive; I don’t think we’ve seen evidence of information leaking through. One of the ways in which we might test for the presence of other universe is … if they were to collide with ours. ..
Nobel Prize for the accelerating universe
Supernova
L: Your work .. and ideas .. around supernova were important. Can we go to the very basics; what is this mysterious supernova?
A: It’s an exploding star. Most stars die a relatively quiet death. Our sun will. A small minority of stars end their lives in a titanic explosion. This is critical to our existence; it is during these explosions that . heavy elements … get ejected into the cosmos … important for ultimately life. We like to study our origins; it turns out these are useful beacons as well. IF you know how powerful an exploding star really is by measuring parent brightness at its peak, in galaxies whose distance we already know—you can calibrate how powerful the thing really is. Then using the known distance of other more distant supernova; you can use their observed brightness compared with their true intrinsic power or luminosity, to judge their distance and hence the distance of the galaxy in which they’re located. Analogy: you judge the distance of an oncoming car at night by looking at how bright its headlights appear to be. Compare it to one that’s a lot closer … We can do that for cars; we can do that for stars!
L: For cars, the headlights, there’s some variation; but they’re somewhat similar. How much variation is there between supernova?
A; There’s several different ways stars can explode. It depends on their mass; whether they’re in a binary system … The ones we use for cosmological purposes, discussing the history of the universe. …. are the so-called Type Ia supernovi. They come from a weird type of a star, called a white dwarf (our own sun will be one in 7 billion years). … it’ll have about half its present mass compressed into a volume just the size of earth—that’s an inordinate density, it’s incredibly dense. .. The matter then is called degenerate matter; the electrons take on a motion (due to Heisenberg’s Uncertainty Principle); meaning a lot of electrons are moving very rapidly, giving the star an extra pressure. So it’s a weird type of star. But normally it wouldn’t explode, and our sun won’t. But if such a white dwarf is in a pair with an otherwise normal star. .. it can steal material from that normal star, until it gets to an unstable limit, roughtly 1.5 times the mass of our sun. This is known as the Chandra Sakar limit … Chandra Sakar was a physicist who figured this out when he was 20 years old; he won the Nobel Prize in physics for it 50 years later, in 1984. His advisor, the great Arthur Eddington, in England, a great astrophysicist; he ridiculed Chandra’s scientific work at a conference in England. … Another inspirational story for the youngster; just keep going; so what if people mock you, even important people! … Anyway, when a white dwarf grows to a certain mass, it becomes unstable. One of the way it can end its life is go through a thermonuclear runaway … Carbon start fusing together … in nuclei. … Doesn’t lead to being dissipated out of the star … … thus you have a runaway, a bomb. Instead of the controlled fusion, which is what our sun does. We need that here on earth, fusion energy to solve our energy crisis. This uncontrolled fusion reaction blows up a star. .. You measure it to be almost the same in all cases. However, the devil’s in the details: we observe them to actually not be all the same. … Variations could be in the amount of trace heavy metals in the white dwarf … is it an old white dwarf or a young one. Part of my work was to show indeed, the 1A’s are different from one another sometimes. You have to calibrate them; they’re not standardized. There is no standard. “STandard candles” is what astronomers like to say. But there are no standard candles; there are standardize-able candles … This is something a colleague of mine, Mark Phillips did; he was on Schmidt’s team and arguably deserved the Nobel Prize. He showed that more powerful Type 1As decline in brightness and rise in brightness as well. If you calibrate this by measuring a bunch of nearby ones. … instead of saying it’s a xy3 supernova; you say, it’s 112 + or - 15 … It tells you where it is in the power scale; it reduces uncertainty. You can then tell .. these things are different, by studying them. … … In 1991 I was skeptical we could use Ias, but then Mark Phillips wrote a paper .. They become calibrate-able and that’s a game changer. Now there are thousands of them … the supernova … project. … Statistical certainties were comparable.
L: How do you feel abut these gigantic explosions … Such explosions could happen in our galaxy? We’d be okay?
A: In most cases we’d be okay; …
L: It gives birth to a bunch of other stars.
A: Yes, and to expanding gases that are … our galaxy is 100,000 light years across … you’d need it to be (this close) to be an existential threat. … potential of certain gases mixing
The greatest story ever told
A: I teach this intro astronomy course at Berkeley; there’s only 5-6 things ONly a few things I’ll want them to remember … I’ll ask this question to them at their deathbeds—I’ll fail them if they don’t know. One question is, where did we come from, the elements in our DNA: the carbon in our cells, the oxygen that we breathe; the calcium in our bones… the iron in our red blood cells, the phosphorus in our DNA; they all came from stars, from nuclear reactions in stars; they were ejected into the cosmos; in some cases like … iron made during the explosions; gases drifted out … mixed with other clouds, made a new star … some of whose elements … enriched the gases in the galaxy progressively more with time. Until finally, 4.5 billion years ago, from one of these enriched clouds, our solar system formed, with a rocky earth-like planet. somewhere, somehow, these self-replicating evolving molecules, bacteria, formed and evolved … paramecia, amoeba, slugs, apes, and us. Here we are, sentient beings that can ask these questions about our origins. With our intellect, with the machines we make, come to a reasonable understanding of our origins. What a beautiful story! If that doesn’t put you in awe of or in love with science and its power of deduction, I don’t know what will! It’s one of the greatest stories … if not the greatest story. That’s a subjective opinion; but perhaps it’s the greatest story ever told. You could link it to the Big Bank … as a subset … it’s an even greater story than the existence of the universe. You could imagine a boring universe that never leads to creatures such as ourselves.
L: Is a supernova usually the intro to that story
A: I touch upon this subject earlier in my course. The sun is fusing hydrogen to form helium nuclei; and some can merge again and form carbon and oxygen nuclei; but that’s where the process will stop for our sun. That’s the beginning of this idea of the birth of the heavy elements; they couldn’t have been born at the time of the Big Bang. That’s the beginning, but you need some of these stars to explode. If they remained forever trapped in the cores of stars, then they wouldn’t be available … for the formation of new stars, solar systems …
Richard Feynman
L: You called RF a mentor of yours when you were at Cal Tech
A: He was quite a character and one of the deepest thinkers of all time, probably. He was the physicist with the most intuitive understanding of how nature works, of anyone I’ve met. I learned a number of things from him, though he was not my thesis advisor. I worked with ____. Feynman I had for two courses. One was general theory of relativity; the other applications of quantum physics. He had this very intuitive way of looking at things, that he tried to bring to his students; he felt that if you can’t explain something in a … simple way, to a non-scientist … Then you probably don’t understand it very well yourself, not very thoroughly. IN me, that created a desire to be able to explain science to the general public. I’ve often found, in explaining things, there’s a certain part I didn’t understand myself. I sometimes find my explanations are lacking in my own mind.
L: If I could pause for a second. You said he had one of the most intuitive understandings; can you break apart what intuitive means.
A: . .How do you draw a picture on paper of what’s going on. He’s famous in this regard for his Feynman diagrams … What you have is an exchange of photons, between charged particles. .. maybe even … There are ways of doing calculations that are brute force. Julian Schweinger … developed math for that. … A set of rules … what to do at the vertex. … Two particles coming out again; … particles splitting off from one another. It looked like a hodge podge at first. They saw you could do these complex calculations in a circular way. Freeman Dyson had a better way of explaining what electrodynamics was … Of the people I knew, Feynman was the most intuitive—is there a picture, a simple way you can understand this. … The path a particle follows: you can get the classical path for a baseball, using physics, if you want. IN a sense, the baseball sniffs out all possible paths. It goes to the Andromeda galaxy …
Meaning of life
A: Life is what you make of it, really. EAch of us has to have our own meaning. It doesn’t have to be … in many cases, meaning is associated with goals, or expectations. You set some for yourself; things you want to do or experience. To the degree that you do or experience those things, it can give you meaning. You don’t have to change the world Newton or Michaelangelo or Da Vinci did; come on, there’s 7.5-8 billion of us now. Most of us aren’t going to change it. Does that mean we’re living meaningful lives? … It just has to give you meaning. Often it’s the things that help others that give us meaning … You taught someone to read; you brought up a nice family; you did a good job; you read a lot of books (if that’s what you wanted), you traveled the world (if that’s what you wanted). If some of these things are not within reach; you’re in a socioeconomic position … you find other forms of meaning. It doesn’t have to be some profound, I’m gonna change the world, I’m gonna be the one that everyone remembers, type thing.
L: In the context of the greatest story ever told; we’re two apes asking about the meaning of life—how does that fit together?
A: It’s a beautiful universe that allows us to come into creation. It’s a way that the universe found of knowing, of understanding itself. I don’t think inanimate rocks, stars, black holes and things have any real capability of abstract thoughts and of learning about the rest of the universe, or even their origins. They’re just a pile of atoms … that has no conscience. …
L: It’s fascinating to think the universe created us; and eventually we built telescopes to look back at it …
A: It’s magnificent! Needn’t have been that way! You can alter the laws of physics, or even the constants of nature—like the mass ratio of the proton and the neutron. Wake me up when it’s over; what can be more boring. You play with it, and you get boring universes: only hydrogen, or only helium. … Even a rich periodic table wouldn’t be possible if certain constants weren’t this way. But they are! That to me, leads to the idea of a multiverse; the dice were thrown many many times. There’s a cosmic archipelago … many universes are very boring. We’re in the rare breed … that is darn interesting. … Just like there are lots of planets; occasionally there are going to be ones … conducive to complexity.
#138: Ayn Rand geek Yaron Brook
Yaron Brook, one of the best-known objectivist philosophers and thinkers in the world. Objectivism is the philosophy developed by Ayn Rand, that she first put forward with her fiction books, and later elaborated on in nonfiction essays and books. Yaron is Chairman of the Board at Ayn Rand Institute and host of the Yaron Brooks Show. He’s co-author of Free Market Revolution: Equal Is Unfair. He analyzes systems of government, human behavior, and the human condition from the perspective of objectivism.
I first read Atlas Shrugged and The Fountainhead early in college. (Along with a lot of other stuff.) I always had an open mind; curious to explore the ideas of thinkers throughout history. No matter how …. dangerous they were considered to be. Ayn Rand was, and still is, a divisive figure. Some people love her; some people dismiss her. I like to look past the flaws of the person, and examine the ideas she presents. I hope that you will be patient and understanding as I venture out across the space of ideas. … Sometimes saying stupid things, but always striving to understand how we can build a better world together.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
Principles of a life well lived
L: The biggest possible question: What are the principles of a life well-lived
Y: I think it’s to live with thought; that is, to live a rational life; to think it through. So many people, in a sense, are zombies out there. .. Their mind is not focused, on what do I need to do to live a great life. Too many people go through the motions … I think the secret to living a great life is to take it seriously; what it means … is to use the one tool that makes us human; … to apply it to living. People apply it to their work … to programming … imagine they used that same focus to actually living life, to choosing values they should pursue.
L: I wear this silly suit and tie; it makes me feel I’m taking the moment really seriously.
Y: That’s right. Each of us has different ways to condition our consciousness …. it’s a conditioning of your consciousness to, now I’m focused, I’m .. … … people don’t have a plan … for their life and how to live it. … To me, reason is this massive evolutionary “achievement.” Any other sophisticated animal; everything has to be coded, written into the hardware. They have to have a solution for every outcome. Human beings have this capacity to self-program. It’s not a tabulo rasa … obviously we have a nature; our brains are structured in a particular way. But given that, we have the ability to turn it on or turn it off. We have the ability to commit suicide; to eject our nature, to work against our interests. So that choice, that fundamental choice—Hamlet says it, to be or not to be. There’s also: to think or not to think; to engage or not to engage. … It requires an act of will to say, okay, I’m awake, I have stuff to do. Some people never awake; they live in that haze. To handle a difficult math problem, you have to turn something in … That is not determined in the sense that you have to focus. You choose to, and you could choose not to.
Free will
L: This is more important than anything else …
Y: I think newer scientists (neuroscientists) pretend they know a lot more about the brain than they really do.
L: Shots fired! I agree with you
Nature of reality
Ayn Rand
Y: …. .. there’s never been a thinker like her. People are afraid of that … she’s basically telling you to re-think almost everything. … She points out to (people) that they’ve lost something. They’ve lost their idealism. They’ve lost their youthful idealism. What makes youthfulness meaningful, other than (being) in better physical shape? When we’re young, some time in the teen years, there’s something that happens to human consciousness. We almost awaken anew.. … We discover this tool, our mind, is available to us, to discover the world, and discover truth. It is a time for idealism! I want to go out there and experiment!
Objectivism
L: Another impossible task … giving a whirlwind overview of the philosophy of objectivism; that of Ayn Rand
Y: She does it in an essay. … She wanted to be … a writer. But she had a particular goal in her writing. … An idealist … She wanted to portray the ideal man. You have to ask that question—what does that mean “ideal man” … you might have glimpses of it in other people’s literature. She reads philosophers about it. What she finds horrifies her. … Nietzsche … has a vision … even though his philosophy is very flawed, but he has that vision of what’s possible; and she’s attracted to that … She’s going to have to develop her own philosophy; she’s going to have to discover these ideas for herself. The glimpses in Aristotle, in Nietzsche, aren’t fully fleshed out. So she develops a philosophy for a very practical purpose—to write, to write a novel about the ideal man. Atlas Shrugged is the manifestation
L: As an aside, when you say man, you mean human. Maybe you can elaborate on how she specifically uses “man” and “he” … we live in a time now …
Y: She did it in a sense, that everybody did it in that period of time. Only in our particular time do people use he/she. But, in this case, in this one sense … she probably meant “man” … A) She viewed there are differences between men and women; we’re not the same. But she was working on a particular vision … she considered herself a man-worshipper. She worshipped the hero in man. She wanted to fully understand what that was. It has massive implications on the ideal woman; and she does portray the ideal woman in Atlas Shrugged. I think her selfish goal, what she wanted to get out of the novel, is the excitement … of seeing your ideal manifest in reality. That which you would be attracted to: fully, sexually, et cetera.
L: So there was no ambiguity of gender
Y: Very much. Remember, the context of Atlas Shrugged portrays a woman running a railroad better than any man could; achieving huge success, more than any man. But even her (character name of Dagny) … needs a man. And that … a character’s name I won’t mention, because it gives away too much of the plot.
L: I like that you do that!
Y: She described herself once as a male chauvanist. She likes the idea of having a man … But more metaphysically she identifies something in the way a man relates to a woman and a woman relates to a man.
L: Let’s not take too far of a tangent … But she was a feminist to me. Perhaps technically you disagree … To me, she had some of the strongest female characters in the history of literature.
Y: … a woman who is sexually, in a sense, assertive.
…
L: To me, as a boy, when I was reading Atlas Shrugged, that was one of the early introductions …. of like a bad-ass lady. I love engineering; I loved it that, here’s a lady that’s running the show. But, objectivism.
Y: She developed it for a novel; she spent the latter part of her life, really articulating the philosophy. She applied it to all these issues. From 1957 until her death in 1982. … So there’s five branches in philosophy. You start with metaphysics; the nature of reality. Objectivism argues reality is what it is. A = a. You can wish it to be b, but wishes do not make something real. Consciousness is there to observe; to give us information about reality; that’s the nature of consciousness. In metaphysics, existence exists. The law of identity; the law of causality; a thing acts based on its nature. Not randomly or arbitarily. We have a tool to know reality; our tool to know reality is reason. Our senses and our capacity to integrate info from our senses. We don’t know the truth from revelation; we don’t know truth from emotions; although emotions are interesting. But they aren’t tools of cognition; they don’t tell us what’s out there in reality. Reasons is a means of knowledge, our means of survival. Only individuals reason, in the same way only individuals can eat. We don’t have a collective stomach. Similarly, we don’t have a collective mind. Only a person can reason. We all think for ourselves; it is our fundamental basic responsibility to live our lives. Once we choose to live, to live to the best of our ability. In morality, she is an egoist. She believes the purpose of morality is to provide oneself with virtues … Happiness is the moral purpose of your life; the purpose of morality is to guide you toward a happy life. She rejects the idea that you should live for other people; your purpose is not to make others happy. But she also rejects the idea that you should use other people for your own purposes. Every person is an end in itself; every person’s moral responsibility is their own happiness. You shouldn’t exploit others, and you shouldn’t allow yourself to be exploited by others. … What is it that produces human happiness? It goes back to our minds, to reason. What does reason require to work effectively? It requires freedom. The enemy of freedom is force; coersion, authority. The Catholic church, doing what they did to Galileo. That restricts his thinking; when he’s in house arrest, he’s not going to discover new truths. It’s too dangerous. Force, coersion are enemies of reason; and what reason needs is to be free. We need to create an environment in which individuals are free to reason. We come up with a concept of individual rights. These define the fact that we should be left alone, free to pursue our values … And that the job of governments is to make sure that we are free. … … she rejects government involved in any aspect
L: The most shallow of questions. The name “objectivism.” Why not individualism? Why branding it that way?
Y: She had a branding meeting. I don’t think objectivism was the first name she came with. The problem was, the other names were taken. For example, rationalism, could have been a good word. Or reason-ism; although that sounds weird. Too many s’s. Rationalism was already a philosophy, inconsistent with hers. Reality-ism, just doesn’t work. Objectivism; I think actually it’s a great name, because it has two aspects to it. Her unique view of what objectivity actually means. In objectivity is the idea of an independent reality. There is truth. And then, there’s the role of consciousness; the role of figuring out the truth. The truth doesn’t just hit you. The truth isn’t just in the thing; you have to discover it. Consciousness applies to (it) …
L: And thereby pulls in the individual in that sense
Y: Only the individual can do it. Now individualism would have made the philosophy too political. She always said, “I’m an advocate of capitalism … because I’m really an advocate for rational egosim. I’m an advocate for rational egoism, because I’m an advocate for reason.” For her it was about reason, and her particular view of reason. She writes a bit about her epistemology; I encourage anyone to read. It’s a tour de force … what it means to discover new discoveries, and to use concepts, and how we use them. She has a theory of concepts that is new and revolutionary, and I think essential for the philosophy of science. The more abstract we get … the easier it is to detach them from truths; the easier it is to be inside our heads. I think what she teaches in the book is how to ground your concepts … in reality. … She saw a lot of parallels between math and concept formation. She was taking private math classes before she died. … She was also interested in neuroscience; she believed that had a lot to tell us about epistemology, and about music, and aesthetics. She believed philosophy ought to integrate all these things. Sadly, we view all these different fields as silos. We become like ants; specialized. She was constantly curious; interested in new discoveries, new ideas.
Godel Incompleteness Theorem
Capitalism
Virtue of selfishness
Win-win
Anarchy
About the pandemic economic damage
Y: Restaurants are difficult to manage; most of them go bankrupt anyway. And they shut them down! Many of them will never open. They estimate 50-60% of restaurants in some places won’t open. These are people’s lives; people’s capital, people’s love. What are they gonna do with their lives now? Are they gonna live off the government? It’s disgusting, and it’s offensive, and it’s unbelievably sad. I care about other people! This idea that objectivists don’t care …
L: Something inspiring about them too …
Y: Anybody who does excellent! I love sports; it’s the one realm where we still value and celebrate excellence. I try to be excellent everywhere in my life. .. We made an effort … to keep restaurants going … The problem is: philosophy drives the world. The response to Covid has been worse than pathetic. And it’s driven by philosophy; by disrespect to science; ignorance and disrespect to significance; disrespect to individual decision-making; government has to decide everything for us. … disrespect of markets. The only place in the planet that’s done well with it are parts of Asia. Taiwan did phenomenally with this; the VP of Taiwan was an epidemiologist. South Korea, Hong Kong, Singapore. Very few deaths. The economy wasn’t shut down in any of those places; there were no lockdowns. The CDC had plans, before this happened, on how to deal—good plans! If you asked people around the world, which country is best-prepared for a pandemic, they said the U.S. Because of the CDC’s plans … And yet all of that went out the window; because people panicked, didn’t think, go back to reason. People were arrogant; refused to use the tools they had at their disposal … It’s simple how you deal with pandemics. You deal with them by testing, by tracing, isolating. And you do it well. And you do it vigorously. You scale up to do it. We have the wealth to do that.
Tribalism and division
L: You’ve talked about Donald Trump. I believe the division should be and can be fixed. We’ve talked about the value of reason. … I’ve noticed that the division shuts down reason. They’ll say, “Is he a Trump supporter or …”
Y: It shuts down after that
L: You as one of the beacons of intellectualism. Quite honestly. How do we bring people together long enough to where we can reason?
Y: (sighs) There’s no easy way out of this. The fact that people have become tribal, and they have. Very tribal. And in the tribe, reason doesn’t matter. It’s all about emotion, about belonging or not-belonging. … It took us decades to get back to tribalism, where we were hundreds of years ago. It took millenia to get out of tribalism. It took the Enlightenment to get us out of it; we’ve been in the Enlightenment for about 250 years—and it’s fading! So what would we need to get out of it. We need self-esteem. People join a tribe, because they don’t trust their own mind; they’re afraid to stand on their own two feet. People need self-esteem. To gain that, they have to have respect for rationality. They have to achieve, and they have to recognize that achievement. To do that … We have schools that teach people to feel … We have them leading six-year-olds discussing politics. What?! They don’t know anything! … They talk about this generation of snowflakes; they can’t hear anything they’re opposed to, because they haven’t learned to use their minds. Boils down to teaching people how to think, and how to care about themselves. … When you have self-esteem, it’s easier to think for yourself. I don’t know how you do that quickly. Part of what I try to do is encourage people to do those things. … Early on, you talked about why I’m not more famous. My following is very small, in the scope of things
L; Yours, and objectivism; could you linger on, why isn’t objectivism more famous?
Y: Because it’s so challenging. It isn’t, to me. After the first shock, when I first encountered it … Once I got it, it was easy. In the sense of, yes, this makes sense. It upends everything, because it says, what my mother taught me is wrong. And what politicians say, left and right, is wrong. … What my teachers are telling me is wrong; what Jesus said is wrong. It’s hard.
L: You talk about politics; most people don’t care. The most powerful thing about objectivism is the … my life, how I revolutionize my life. That seems like an important, and appealing … get your shit together!
Objectivism and Jordan Peterson on personal responsibility
Y: This is why Jordan Peterson is more … appealing. His personal responsibility is shallow. Make your bed! He says, embrace Christianity, Christianity is fine. Just do these few things. … And by the way, he says, happiness, you either have it or you don’t. He’s giving people an easy out. … People like easy outs. People buy self-help books.. .. I’m telling people, think! Stand on your own two feet! Be independent! Do your own thing, but thoughtfully, not based on emotions.
L: So you’re responsible not just for a few habits. You’re responsible for ecerything.
Y: Yes, and here’s the big one: You’re responsible for shaping your own soul. You get to decide what your consciousness is going to be like. … Your emotions are a tool. The tools you have is thinking, experiencing, living … listening to great music, watching great movies. Art is very important in shaping your own soul. But it’s work! And it’s lonely work. Because it’s work you do with yourself. If you find somebody who you love who shares these values, that’s great. It’s hard, it’s challenging, it ends your world. The reward is unbelievable. … Think about the Enlightenment. Up until the Enlightenment, where did truth come from? It came from a book. Life sucks, you have nothing, you don’t enjoy it. It’s easy. The Enlightenment comes around and says, we’ve got this tool: it’s called reason! It’s not in a book. Your reason allows you to discover stuff about the world.. … Newton teaches us the laws of mechanics. How does stuff work? … I can use my mind; I can discover truth! Isn’t that amazing!? … How come I can’t decide who I marry? What profession I’m in? … It’s all reasoning. Once you understand the efficacy of your own mind to discover truth, everything opens up. Now you can take responsibility for your own life. But postmodernism tells us there is no truth, there is no reality, and it’s all useless anyway. Critical race theory tells us, reason doesn’t matter; it’s all shaped by your genes and the colors of your skin. You’ve got our friend at UC Irvine telling them, your senses don’t tell you anything about reality … what use is that? It’s complete fantasy. You’ve got every intellectual voice in the culture telling them their reason is impotent. There’s Steven Pinker—who I love, but he needs to be stronger about this. There’s a few people on the IDW who are big on reason, but not consistent enough and without full understanding of what it implies. And then there’s little old me! I’m not only willing to articulate the case for reason … There’s other IDW people; get to reason. All politics, you can be whatever. No! You can’t be a socialist and be for reason. And you can’t be a determinist and be for reason. The whole point of reason is it’s an achievement, and … it requires choice. So it does feel like … little old me. The allies I have; a few libertarians on economics … a few in the IDW … But not in all things
L: For the few folks and thinking about this and the trajectory of their own life, I guess the takeaway is, reason is a difficult project, but a project that’s worthy of taking on.
Y: I don’t know if “difficult” is the right word. It’s difficult in the sense that it requires energy and focus. But it’s immediately rewarding. It’s fun to do. Rewards are immediate. Pretty quick!
#139: Andrew Huberman’s first visit
Andrew Huberman, a neuroscientist at Stanford working to understand how the brain works, how it can change through experience, and how to repair brain circuits damaged by injury or disease. He has a great Instagram account @HubermanLab, where he teaches the world about the brain and the human mind. … Also he’s an inspiration; he shows you can be humble, giving, and still succeed in the science world.
As a side note; let me say, I heard from many people about the previous conversation I had with Yuran Brook. Some people hated it! I misspoke during some parts … was uneducated about some important things that happened in the past. I bring all that up to say that if we are to have difficult conversations, we have to give each other space to make mistakes, to learn, to grow. Taking one or two statements from a three-hour podcast, and suggesting that they encapsulate who I am, I was, or ever will be, is a standard that we can’t hold each other to. The conversation with Yuran is mild, relative to some conversations that I will likely have in the coming year. Please continue to challenge me, but please try to do so with love and with patience. I promise to work my ass off to improve.
<iframe width="560" height="315" src="
title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
L: You’ve mentioned that in your lab at Stanford, you induce stress by putting people in VR and putting them through a set of experiences that scare them. You mentioned this with Joe, or Whitney. What are people afraid? What are these fear experiences that you find to be effective.
A: We should probably define fear, without going too far down the rabbit-hole. You can’t really have fear without stress, but you could have stress without fear. .. You can’t have trauma without fear and stress; but you can have fear and stress without trauma. That is one of the motivations for having a laboratory that studies these things. We need better physiological, neuroscientific, and operational definitions of what these things are. The field of understanding emotions and states … is very complicated. We can do away with a lot of complicated debate, and say in our laboratory, what we’re looking for, to assign it a value of fear, is a big inflection in autonomic arousal—increases in heart rate, breathing, perspiration … all the hallmark, signature features of a stress response. In some cases, we’ve got a benefit of having neurosurgery patients, where we’ve got electrodes in their (brain); we’re getting multi-unit signals; we can start seeing some central features, meaning within the brain. As trivial as it might seem when listening to it, almost everybody responds to height and falling from a high virtual place with a very strong stress, if not fear response. … Because the stimulus, and how it connects to semicircular … inner ear; all that pulls all your physiology … suddenly you’re sweating; you feel as if you’re falling because of the optic flow. We’ve got a dive-with-great-white-sharks experience. We got 360 video from the real world; this was important to us. When we decided to set up this platform; a lot of the studies of these things in laboratories; these people study fear by showing subjects a picture of a bloody arm, or a snake. IT wasn’t creating a real enough experience. We needed to do something where people weren’t going to get injured … people momentarily, forgetting they’re in a laboratory. Heights will always do it. If people want to challenge me on this, I point them to that movie Free Solo. That movie’s popularity—you knew he was going to live when you walked into the theater; and yet it was still scary; people were somehow able to put themselves into Alex’s experience enough; they were concerned … People who have generalized anxiety; they wake up and move through life … we can tip them a little bit more easy with things that don’t necessarily get everyone afraid. Claustrophobia, public speaking. IF you’re afraid of sharks, like my sister—she won’t even come to my laboratory! That’s how terrified some people are of these specific stimuli. … We have you step off a virtual platform. It’s a flat floor in my lab.
L: You actually allow them the possibility in the virtual world to take a leap of faith?
A: We have them play a cognitive game; it gets increasingly complex; it speeds up on them. There’s a failure point for everybody, where you can’t make the commands fast enough. All of a sudden, they’re on a narrow platform between two buildings. We encourage them, to continue across that platform to continue the game. Some of the people will get down on the ground and hold onto a virtual beam that doesn’t even exist. The power of the brain to enter these virtual states that aren’t even real …
Virtual reality
L: We did this with autonomous vehicles; to try to create the experience of a car that could run you over. I was surprised how real that whole world was. I was still afraid of being hit by a car.
A: Even though it was all a simulation.
L: Yeah. It wasn’t even ultra-realistic simulation.
A: Any kind of depth; we’re just programmed, to not necessarily recoil, but be cautious about that deaths. And then looms; anything coming closer. … Some looming-sensing processing in the retina. The way … folks learn to not get eaten by great white sharks. When they start coming toward you, you swim toward them. There’s something about forward movement toward any creature; that creature questions whether it would be a good idea, to move toward you.
L: Critical idea for autonomous driving research … 360 video; you can’t change the reality that you’re watching. Is there something fundamental about fear and stress that the interactive element is essential for?
A; It works best to use mixed reality. We have a snake stimulus; a spider stimulus. I don’t like snakes; the slithering creates a visceral response for me. Some people, less so. One way we can get them to feel more of that; get them to use mixed reality. Use an actual physical bat; get them to stomp out the snake. That tends to be not as stressful as when they use a physical weapon. There’s something about engaging that makes it more of a threat. We always get the subjective report from the subject. We don’t want to project our own ideas of what they were feeling. …
Claustrophobia
Skydiving
Overcoming fears
Optimal performance
Deep work
Psychedelics
Deep work
Everything in the brain is an abstraction
A: … there are beautiful experiments done by Greg Reckinzone up at UC Davis. This will become obvious as I say it; obviously the ventriloquist doesn’t throw their voice. An auditory-visual mismatch … freaks people out. People will recoil. Sounds aren’t getting thrown across the room; this is the way the brain creates these internal representations. … I’m sure the listeners appreciate this, but everything in the brain is an abstraction. The sensory apparti … the interroception, they’re taking signals from inside the body. Taking all that, and the brain is abstracting that; in the same way that, if I took a picture of your face and handed it to you, you’d say, yeah, that’s me. If I were an abstract artist … I might mix it up; I’d make your eyes like water bottles; I’d start assigning fruits and objects to the bits of your face. That’s what the brain does. The neurons … have no resemblance to your face. I’m not sure I’ve fully internalized that. It’s weird to think about; all neurons can do is fire in space and in time … The action potential for a given neuron doesn’t always have the exact same wave form. You can modify that wave.
L: There’s a lot of fascinating stuff with neuroscience; the fuzziness about the transfer of information from neuron to neuron. We touch on it when we compare artificial and biological NNs. The circuitry you’re getting at; the brain, a bunch of stuff firing, forms abstractions … that are beautiful: layers upon layers. When you’re programming … in Python; it’s awe-inspiring, to know it all becomes 0’s and 1’s. In the same way, with the brain; is there something interesting or fundamental to you about the circuitry of the brain … How much do we understand? Focusing on the vision system; is there something specific about the structure and circuitry of the vision system? Or is it all a chaotic mess we don’t understand?
Human vision system
A: The beauty of the visual system; the reason it’s such a great model for addressing these kinds of questions, is we can control the stimuli. Spatial frequency, temporal frequency—how fast things are moving. So many things we can do in a controlled way. Whereas, cognitive encoding, encoding the space of concepts … I like you, am drawn to the big concepts in neuroscience. I confess, in part because I’m not smart enough to go after the really high-level stuff; I always like to address things that are tractable, where we stand to gain ground in a given time. I’m happy to have a talk about consciousness; most don’t want to hear what I have to say. We can save that for later.
L: We talked about psychedelics. Can experiments in neuroscience be constructed to shed any kind of light on these concepts? To me, vision is one of the most beautiful things about human beings. On the AI side, computer vision has some of the most exciting applications of computer networks. It feels like that’s a neighbor of cognition and consciousness.
A: Yeah, the visual system is amazing. To survive, humans mainly rely on vision. It’s a filter for cognition; it’s a strong driver of cognition. We’re moving to high-level concepts … the way the vision system works …
L: Let’s go there
A: The retina, a structure at the back of your eye, as thick as a credit card. IT is a piece of your brain; it’s a forebrain structure; in the first trimester, there’s a genetic program that made sure that that neural retina, part of your CNS, was squeezed out into the embryonic eye cups, and that\ the bone formed with a little hole with optic nerve is connected to the rest of the brain. That window into the world is the only window into the world for a mammal with a thick skull … … Humans have to do all that through the eyes. Three layers of neurons, that are a piece of your brain .. the optic nerve connects to the rest of the brain. .. The central circadian clock tells every cell in your body about time of day. All sorts of good things happen … Good incentive for keeping a relatively normal structure of light exposure.
L: You lost me
A: When you’re young ,it’s easy to go off schedule and recover … We were designed to get light and dark at different times of the circadian cycle. All that info is coming in through specialized information in the retina … That’s not spatial information; it’s subconscious. You don’t think, “Oh, it’s daytime.” IT’s literally counting photons. Even if it’s a cloudy day, lots of photons coming in … That’s the circadian clock. There’s other neurons in the eye; that signal to the brain. They signal the presence of things that are lighter than background or darker than background. It’s looking at pixels; they look at circles. Those neurons have receptive fields. They respond to little circles of dark light … IT sounds very basic, it’s like red-green-blue, circles lighter or dimmer than what’s next to it—that’s the only information sent down the optic nerve. … Spikes, action potentials, in space and time. It’s beautiful to think of how all that information from the outside world … is converted into a totally different language like morse code. The thalamus … like the retina does. The thalamus is basically weighting things. … That signal, I’m going to allow up to the cortex. That signal is more like the red next to it—throw that out. Cortex of course is where perceptions happen. V1, visual area 1, you start getting representations of things like oriented lines. … That’s how the visual system is organized all the way up to the cortex. It’s hierarchical; you don’t build up that line by suddenly having a neuron that responds to lines in a random way. … It’s not random. There’s no abstraction at that point, in fact. If I showed you a black line; I could be sure if I were imaging V1; I would see that representation of that line somewhere in your cortex. It’s concrete; it’s not abstract. But then things get mysterious. Some of that info goes further up into the cortex … Nancy Kenwood at MIT … talks about this … face area. You find single neurons that respond only to your father’s face; or to Joe Rogan’s face. The orientation of his face; he’s represented in some abstract way by a neuron; the Joe Rogan neuron.
L: It might have limits; like I might not recognize him if he were upside down
Neuralink
A: Nancy’s lab has done that. She found there are space neurons. It turns out, she was right; they use stimuli called Griebel stimuli, which morphs a face into something gradually that looks like an alien thing they call a Griebel. The neurons don’t respond to Griebels. … Forgive me (for oversimplifying this summary) … Visual Area 4 has color pinwheels. .. The stimuli get more elaborate; at some point, you start getting abstract representations that can’t be explained by simple point-to-point wiring … What we’re talking about maps to the auditory system. … A Doppler, hearing a car passing by. At some point, you get into motifs of music. You start abstracting. If you start thinking about concepts of creativity, and love, and memory. What is the map of memory scripts? Presumably, there’s enough structure at the early stages of processing .. that you have the building blocks, your 0’s and 1’s, but you depart from that eventually. The exception to this—I was just mainly talking about the neocortex (with its six layers), is that subcortical structures are more like machines … That which controls breathing … neurons that respond to things like temperature on top of my left hand. I came into this from a perspective of psychology. I forced myself to learn some molecular biology—one of the most beautiful experiences you can have in life, is to lower an electrode into the cortex … show a stimulus; you can convert the recordings coming off that electrode. You hear the “hash,” not the hash you smoke. In the cortex, you eventually find the stimulus … It’s very (loud) … When you drop ___ into the thalamus, it’s like a machine; you never hear hash. You drop the electrode down; it doesn’t have to be a high-fidelity electrode; you plug it into a monitor, it’s picking up electricity. … If you walk in front of that animal or person … You could do that for two days, and that neuron, every time there’s a stimulus, it fires. … When you go sub-cortical … there’s no abstractions. It’s 2 + 2 = 4. I know we have some common friends at Neuralink; I’m a huge fan of what they’re doing and where they’re headed. But my question was, to some of them, when are you going to go sub-cortical … In those areas, a stimulating electron can evoke a behavior or a state. … In figuring out how the brain works … … .Necessary, but not sufficient. … Eventually, we’ll get there. If you’re talking sub-cortical circuits, that’s where the action is. You could potentially cure Parkinsons’s … For those .. interested in neuroscience … … Let’s restore motion to the Parkinsonian … let’s restore speech to the stroke patient. There’s nothing abstract about that. So anyway … I’ve mixed in a lot of opinion there. But having spent some time, like 25 years, digging around in the brain, I think given it’s 2020, we need to ask better questions. The really high-level stuff is fun, it makes for good conversation; it’s brought enormous interest. The stuff about consciousness and dreaming; I don’t know that we’re there yet.
L: You think there might be a chasm between the two views: the power of the brain arising from … the circuitry that forms abstractions … or the majority of the circuitry that’s doing brute-force dumb things …
A: … ON your podcast, Elon Musk said the brain is a monkey brain with a supercomputer on top. Great description; it captures a lot of important features. Limbic friction, right … think about the drug addict driven to go pursue heroin or cocaine. .. They use their cortex; it’s been hijacked by the limbic system … Not fair to monkeys; monkeys can make plans.
L: You’re saying there’s a lot of value to focusing on the monkey brain
A: I do .. if I could put a chip anywhere in the brain today: I’m not sure I’d put it in the neocortex. The reason is it’s an abstraction machine … it’s possible that our abstractions are so different, I wouldn’t know what patterns were firing. But if I wanted to increase my level of focus and creativity, I’d want to control my own level of limbic friction. Or ramp up pre-frontal cortex’s activation. A lot can happen in the thalamus; you can allow more creative thinking by lateral connections. These are experiments I’d want to do … You can look at what abstract behaviors would arise from that. There’s this obsession with the neocortex. I might lose a few friends; hopefully I’ll gain a few also. One reason people spend so much time in the neocortex; I have a fact and an opinion. One fact is, you can image there, and you can record there. Some microscopy that allows you to image deep into the brain, still don’t allow you to image very deep … Still, it’s very narrow; it’s like looking at the bottom of the ocean, through a spotlight. It’s easier to look at the waves up on top. .. Let’s face it; one of the reasons, there’s so many recordings there, is it’s very hard to image deeper. Now, the microscopes are getting better … All this advanced microscopy; it’s hard to image deeper. .. Now the other thing, which is purely opinion. As long as there’s no clear right answer; it’s easier to do creative work in a structure which, no one knows how it works. If you’re working in hypothalamus, structures we knew about since 60s and 70s; you have to combat existing models. Whereas in cortex, neocortex, no one knows how the thing works. There’s a lot of room for discovery. I think with the tools that are available nowadays, and where people are … We have to be careful and thoughtful about what are we trying to write? There are many brain structures for which we already know what scripts they write. I don’t think it’s boring—the fact they act like machines makes them predictable …. The neocortical jockeys out there; they … I should call out experiments done by (Japanese guy) from MIT; as well as Mark Mayford’s lab at UC San Diego … they monitored neurons as an animal learned something. .. You monitor the neurons … Neurons 1-100 were played in a particular sequence. The notes that gave rise to the song that led to the behavior. You play the keys, you get the same behavior. So the space-time code may be meaningless for some structures. That’s freaky, a scary thing. IT means all the space-time firing in the cortex; the space part may matter more than the time part. ..
L: You’re saying some of the magic is in the early stages; closer to the raw information
A: I think so. … sensory transformations. I can show you a red circle; look at how many times the neuron fires in response to it. That is the transformation; you converted red circle into three action potentials. Beep-beep-beep.
Science of consciousness
David Goggins
A: Some people will reward the pain process so much that friction becomes the reward. People like Goggins … who have gone through cancer treatment three times … the Victor Frankl stories, Nelson Mandela—I’m sure the same process is involved. This speaks to the generalizablility of these processes … Goggins talks about eating souls. In his mind, that’s a form of reward. IT’s not just picking up a trophy or something; it actually gives more energy, more neural energy, more dopamine …
L: So it ultimately maps to that. Yeah, he creates enemies. .. I never think I have enemies; they’re versions of me inside my head. Through that 30-day challenge I tried to come up with fake enemies; but it didn’t work
A: You certainly have a formidable adversary in this one …. Let’s hope you both survive this one
L: My problem is the physical .. everything we’ve been talking about the mind. There’s a physical aspect—when you injure yourself, you just can’t function.
A: You’re talking about taking yourself out of running
L: … for the rest of your life … I’d like to avoid that. Keep it purely in the mental. The problem with these physical challenges, as David has experienced—it has a toll on your body. I tend to think the mind is limitless; the body is limited.
A: Key is to dynamically control your output. By reducing effort, and restoring by subjective reward processes … bring resources to animal or person through foraging. They work so well, because they’re down in those circuits where we know the 0’s and 1’s … That’s great; it can be subjective in that “I reached this milestone.” If you do self-reward it, it’s effort minus one …
Science communication
L: I have to ask … You’re one of the great communicators in science. Do you have a philosophy behind it; is it just an instinct? I’m excited that, you know, somebody from Stanford—so I’m in multiple places, in the sense of where my interests lie. Politically speaking, academic institutions are under fire, for many reason we don’t need to get into. But I believe in places like Stanford and MIT; as one of the most magical institutions for inspiring … people to dream, build the future. I believe it is a really special, these universities are really special places. It’s always exciting to me when somebody as inspiring you represents those places … it makes me proud, that somebody from Stanford … somebody like you is representing Stanford. How did you come to be who you are, in being a communicator?
A: First of all, thanks for the kind words. Stanford is an amazing place, as is MIT. I’ve got many friends at MIT.
L: Stanford’s better, BTW. Smarter friends. …
A: i think the great benefit of being at a place like … you look around and the average is very high. You have many best-in-class; top two or three in the world at what they do. It’s a great privilege to be there. … there’s an emphasis on what gets exported out of the university. Trying to keep an eye on what’s needed in the world, trying to do something useful. Proximity to … the Silicon Valley … So the reason I got involved in educating on social media was actually because, Pat Dosset, the bear crawl guy … we had formed a good friendship; he talked me into doing early-morning cold-water swims. We were talking one morning: he said, “What are you going to do to serve the world in 2019? What are you gonna do that’s new?” I said, if I had my way, I’d teach people, everyone, about the brain. He said, “do it. Shake on it.” I started putting out these posts. … You asked about a governing philosophy. I want to increase interest in the brain in biology … I’d like to increase scientific literacy; it can’t be rammed down people’s throats. Statistics, genetics … it has to be done gradually in my opinion. I want to put valuable tools into the world … that map to things we’re doing in our lab. … :Reduce stress; raise one’s stress threshold … not just being calm but how to tolerate being not-calm. There’s a ton of micro-missions in this. IT all maps back to the 8- and 10-year-old version of me. I used to read about medieval weapons; and I would ask if I could talk about it to the class, and teach. I promise, I don’t really like being the point of focus. I just get so excited about these gems I find in the world, in books, experiments, and discussions with colleagues. I just compulsively have got to talk to people about it. I try to package it into a form that people can access. .. Stanford has been very supportive, thankfully. They’ve re-posted some stuff on social media. It’s a precarious place to put yourself out there as a research academic. … Some probably wonder if I’m still serious about research, which I absolutely am. I acknowledge that their research … needs to be talked about. I don’t like the phrase, “dumb it down.” What I like to do is take a concept which people will find interesting … If someone’s visiting your home, you’re not going to cram foise gras down their face. … Rather, you’d slowly walk them toward it. The best information prompts more questions of interest. One door opens, then another door opens; the image in my mind, you create a bunch of neuroscientists who are thinking about themselves neuro-scientifically. .. I cast a lens onto what I think are a bunch of interesting topics. Some day, I’m going to go into the ground. I’m comfortable not everyone’s going to be happy with how I deliver the information.
Man's Search for Meaning
L: You mentioned Victor Frankl; I re-read that book quite often. What do you think is the meaning of it all? Do you mention that book from a psychologist’s perspective, which Frankl was? Or do you think about big philosophical questions it raises?
A: One of the great challenges … is illustrated best by the Frankl example. Our sense of meaning is very elastic in time and space. We talked about this earlier; it’s amazing to me that someone locked in a cell or concentration camp can bring the horizon in close enough … they can micro-slice their environment … even in a little square box, or terrible situation … Speaks to one of the most important features of the human mind. Take two opposite extremes. Let’s say an alarm went off in this building, and it started shaking. OUr vision, our hearing, … Everything would be tuned to this space-time bubble. .. the only meaning would be, get out of here safe, contact loved ones, et cetera. … IF we were to sit back totally relaxed, we could do the Pale Blue Dot thing … where we imagine ourselves in this room, in the U.S., in the Earth … it can seem so big, that all of a sudden, it’s meaningless. If you see yourself as just one brief glimmer in all of space-time. “I don’t matter.” And if you go to … this text thread on Instagram, everything seems inflated, and the brain will contract and dilate its space-time vision and time, but also sense of meaning. That’s beautiful, it allows us to be so dynamic in different environments; we can pull from the past and the future. That’s why examples like Mandela and Frankl … had to include … dopamine rewards in the boxes they were forced into. I’m not trying to dodge an answer. For me, personally … I have this complicated history in science … where all my advisors died young. They were wonderful people and had immense importance in my life. What I realized is, we can get so fixated on the thing we’re experiencing as holding tremendous meaning … it only holds that meaning if we’re in that space-time regime. What really gives meaning is that you can move between these different space-time dimensionalities … I’m not trying to sound like a theoretical physicist. … What I want to do in this lifetime is, engage in as many different levels of contraction and dilation of meaning as possible … If I pulled over to the side of the road, I bet I’d find an anthill, and the whole world is fascinating. The key is the journey back-and-forth, up-and-down that staircase. My goal is to get as many trips up and down as possible before the Reaper comes for me.
L: Wow, beautiful!
A: I watched people die; my post-doc advisor. But they found beauty in these closing moments. There bubble was their kids in one case; one was a Giants fan; she got to see a Giants game. Learning how to do that better and more fluidly; recognizing where one is. Not getting too attached to the idea that there’s one correct answer.